Comparing the effects of ECT and Intravenous Ketamine in psychiatric patients with major depressive episodes

Abstract

Background:

Electroconvulsive therapy (ECT) and intravenous (IV) ketamine are treatments used for severe depression and/or treatment-resistant depression (TRD). ECT is considered one of the most effective treatments for severe depression, although there is debate within the field regarding the effectiveness between ECT and IV ketamine in comparable TRD patient groups.

Methods:

This retrospective chart review of open-label, nonrandomized treatment from a psychiatric hospital compared the effects of up to three weeks of ECT and IV ketamine in patients with major depressive episodes (MDEs). Our cohort included 146 inpatients and outpatients aged 18 to 74 years old. 94 patients received subanesthetic IV ketamine infusions 2 times a week, and 52 patients received ECT treatment 2-3 times a week. The primary outcome measure was the Montgomery-Asberg Depression Rating Scale (MADRS).

Results:

Overall on the MADRS, 45.2% of participants showed clinical symptom change between the beginning of treatment compared to the end of treatment, while 54.8% did not. Reliable change index (RCI) analysis indicated 58.9% showed significant symptom change, while 41.1% did not. Chi-Square tests revealed significant associations between treatment type and clinical symptom change on the MADRS, with patients who received ECT showing greater symptom improvement compared to those who received ketamine (p < 0.05). Response rates of ECT and ketamine were 67.3% and 45.7%, respectively, whereas remission rates for ECT and ketamine were 60.0% and 46.1%, respectively.

Conclusions:

ECT demonstrated the more robust antidepressant effects compared to ketamine. Randomized comparative trials are required to gain a better understanding of these modalities.

1. Introduction

Major depressive disorder (MDD) is a serious public health concern. In 2023, 21.9 million adults in the United States reported experiencing at least one major depressive episode (MDE) (American Psychiatric Association, 2013; Substance Abuse and Mental Health Services Administration, 2023). While evidence-based treatment options for individuals with MDD and BD exist, such as pharmacology (e.g., SSRIs, atypical antidepressants) or psychotherapy (e.g., cognitive behavioral therapy), approximately 30% do not respond to first-line treatment methods and are considered “treatment-resistant” (McLachlan, 2018). Treatment-resistant depression (TRD) is broadly defined as those who fail to respond to two or more antidepressant trials (Al-Harbi, 2012; Elsayed et al., 2022; Gaynes et al., 2020). Individuals with TRD are particularly vulnerable, experiencing higher healthcare costs and being at greater risk of suicide (Brenner et al., 2021; Zhdanava et al., 2021). Consequently, it is imperative to investigate additional treatment modalities for those with unresolved MDEs.

For patients suffering from TRD, various strategies exist, including combination pharmacotherapy and psychotherapy, as well as neuromodulation interventions. Of the latter, ECT remains the best studied. The ECT device is FDA approved for use in patients with unipolar or bipolar depression among other disorders (Espinoza and Kellner, 2022). However, deciding which strategy is most likely to help in TRD remains trial and error.

Over the last decade, intravenous (IV) ketamine infusion, often delivered over a series, although not FDA approved for use in MDD or TRD, has gained popularity as an adjunctive treatment for depression. During treatment, ketamine, an antagonist of N-methyl-D-aspartate receptors, is administered to patients intravenously in subanesthetic doses 2-3 times per week over several weeks, followed by maintenance sessions at a decreased frequency. Studies have shown that IV ketamine may mitigate depressive symptoms, with patients experiencing relief after a single session, although the effects are often transient (Kishimoto et al., 2016; Romeo et al., 2015; Rybakowski et al., 2017). Despite not currently being FDA approved for depression, ketamine use has expanded to clinical settings as a result of a studies demonstrating relatively safe use over short periods of time, generally good tolerability in most patients, and faster onset of antidepressant effect (Yavi et al., 2022).

ECT is established as one of the most effective treatments for severe depression, with studies showing pooled response and remission rates ranging from 70-90% (Espinoza and Kellner, 2022; Kellner et al., 2020). ECT involves delivery of a brief electrical stimulus to the brain to trigger a short seizure while the patient is under general anesthesia. An ECT treatment course comprises between 6-12 sessions over 2-4 weeks to produce the best and most sustained response (Espinoza and Kellner, 2022). Although ECT is the most effective neuromodulation technique (Gazdag and Ungvari, 2019; Karlovic et al., 2020), it remains underutilized due to misunderstanding and controversy among providers and patients. Misconceptions about ECT, including its mechanism of action, the role of induced seizures, and the fears about memory loss, are common among both patients and providers. Although concerns about ECT persist, particularly regarding its association with cognitive side effects (Park et al., 2021), many concerns are also shaped by outdated or dramatized portrayals of ECT in the media (McDonald and Walter, 2009).

IV ketamine has gained in popularity as an option for TRD due to its purpoted advantages in cognitive tolerability and ease of use. Although IV ketamine has been shown to decrease depressive symptoms, there is debate about whether IV ketamine is as effective as ECT across all levels of depressive severity, ages, settings, and associated symptoms. Research has produced incongruous findings when comparing the effectiveness between the two techniques. While a large body of studies support the view of ECT’s superiority compared to ketamine (Folkerts et al., 1997; Petrucci et al., 2024; Rhee et al., 2022), other studies have found no differences in the rates of clinical improvement between ECT and ketamine (Kheirabadi et al., 2019). Importantly, a recent open-label, randomized trial found that ketamine was non-inferior to ECT in TRD without psychosis (Anand et al., 2023). However, another noninferiority trial in a different setting and patient population found that ECT was superior (Ekstrand et al., 2022). A secondary analysis found that ketamine may lead to greater improvement in depressive symptoms compared to ECT, which contradicts previous findings that established ECT as superior to ketamine (Jha et al., 2024). Here, we conducted a retrospective chart review of open-label, nonrandomized treatment comparing the effectiveness of ketamine and ECT in a cohort of patients with MDEs. We hypothesized that patients who received ECT would have greater improvement in depressive symptoms compared to those who received ketamine treatment.

2. Methods

2.1. Demographics

The IRB approved chart review (Protocol H-52110, Baylor College of Medicine IRB) identified 204 inpatients and outpatients between the ages of 18 to 74 years old who received either ECT (n = 63) or ketamine (n = 141) treatment at The Menninger Clinic between September 1, 2019, and December 30, 2022. Note that we excluded 58 patients from the analysis due to missing outcome measures after the baseline. Therefore, a total of 146 patients (ECT: n = 52, ketamine: n = 94) were included in this study. All relevant patient information, including age, sex, diagnoses, medications, and clinical outcomes, was extracted from each patient’s chart (Table 1). Of note, medication data reported in this study reflect the medication regime that patients were prescribed at the start of their treatment. Medication regimens were not standardized or restricted as part of our study protocol. Medication adjustments could occur based on their psychiatrist’s clinical judgment, although any such changes were not controlled for in our analysis. All patients included in this sample were required to have the capacity to provide informed written consent prior to their treatments.

Table 1.

Demographic Characteristics, Medication Regime, and Psychiatric Diagnoses Across Groups

column1d	ECT (n = 52)	Ketamine (n = 94)
Age (Mean ± SD)	46.23 ± 15.26	36.56 ± 16.2
Female sex, n (%)	30 (57.7%)	55 (58.5%)
White race, n (%)	46 (88.46%)	85 (90.4%)
Medication Regimen
Antidepressants	41 (78.8)	79 (84)
Antiepileptics	15 (28.8)	34 (36.2)
Antipsychotics	38 (73.1)	47 (50.5)
Benzodiazepines	30 (57.7)	28 (29.8)
Hypnotics	4 (7.7)	7 (7.4)
Stimulants	~	6 (6.4)
Opioids/Non-opioids	1 (1.9%)	4 (4.3)
Psychiatric Diagnoses, n (%)
MDD	35 (67.3)	53 (56.4)
PDD	9 (17.3)	27 (28.7)
Unspecified Depression	2 (3.8)	4 (4.3)
Bipolar I	~	6 (6.4)
Bipolar II	4 (7.7)	2 (2.1)
Other Specified Bipolar	1 (1.9)	~
Unspecified Bipolar	1 (1.9)	2 (2.1)

ECT: electroconvulsive therapy, SD: standard deviation, MDD: major depressive disorder, PDD: persistent depressive disorder

2.2. Treatment

Before beginning any treatment, patients received an initial consultation to assess their current psychiatric status and medical concerns, helping to determine the most appropriate treatment option based on their symptoms and medical history. Note that there was no standarized assessment protocol or decision for treatment selection, but treatment decisions were guided by exclusion criteria and patient preference. All patients included in this sample were not formally assigned to either treatment method, and selected treatment was based on individual patient needs and shared decision-making between the psychiatrist and patient. Both inpatients and outpatients were eligible for treatments. To determine ECT eligibility, relative contraindications (e.g., recent myocardial infarction, brain tumors, stroke, chronic lung diseases) were considered. To determine ketamine eligibility, patients were excluded if they have current or past psychosis and if they had substance use disorder except for cannabis. Eligibility was also determined by other factors including cost, insurance coverage, and transporation. Prior to each session, patients were provided informed consent and completed a brief intake to review medical and psychiatric status.

2.2.1. ECT Procedure

Standard monitoring was used throughout the procedure, including electrocardiogram (EKG), electroencephalogram (EEG), blood pressure, oxygen saturation, pulse oximetry, and end tidal carbon dioxide monitoring. ECT was administered using a Thymatrom System IV machine. Treatments occurred 2-3 times per week. Right unilateral (RUL) and bitemporal (BT) electrode placements were used. RUL treatments were delivered at a 0.3ms pulse width and BT treatments were delivered at 0.5-1ms pulse widths. Electrical dose was based on empirical dose titration to determine seizure threshold. Patients who started with RUL electrode placement but did not have a 50% reduction in symptoms after 6 treatments were switched to BT treatment. If electrode placement was switched, another empirical dose titration procedure was conducted. The details of the ECT procedure can be found in Supplementary Material.

2.2.2. Ketamine Procedure

Standard monitoring, including blood pressure and pulse oximetry, was used throughout. Patients started with 0.5mg/kg ketamine infused over 45 minutes. If MADRS scores did not decrease by 50% after two sessions, the dose was increased to 0.75mg/kg, and then to 1mg/kg under the same criteria. For patients with a body mass index (BMI) over 30, dosing was based on adjusted body weights. Adjunctive medications for heart rate, blood pressure, and nausea were given as needed.

2.3. Clinical Outcomes

The Montgomery-Asberg Depression Rating Scale (MADRS) (Montgomery and Asberg, 1979) was used as our primary outcome measure to assess changes in depressive symptoms from baseline (T1) to at the end of the initial 3-week period (i.e., post-treatment. T2) in patients who received either ECT or ketamine treatment. Note that the MADRS was measured before the treatment (i.e., T1) and weekly thereafter, with the last MADRS measurement used as the post-treatment time point (i.e., T2). The MADRS was administered by mental health professionals. Assessments were not blinded. That is, neither the patients nor the raters were blind to the treatment condition. We also calculated response rates (i.e., a decrease of ≥ 50% from baseline) and remission rates (i.e., MADRS score < 10 at post-treatment) to each treatment. No formal self-reported patient outcome measures or systematic side-effect data were included in this retrospective chart review.

2.4. Data Analysis

Clinically meaningful change was defined using (Jacobson and Truax, 1991) reliable change index (RCI) and clinical cut-off methodology. The RCI is a statistical method used to determine if the change in the individual’s score (i.e., clinical outcomes) is statistically significant (RCI ≥ 1.96), by dividing the change in the individual’s scores by the standard error of the difference for the test (Jacobson and Truax, 1991).

We also calculated the clinical cut-off score at which the probability of a score belonging to either a clinical or non-clinical population is equal, following the method proposed (Jacobson and Truax, 1991). To determine this threshold for the MADRS, we used the current sample’s mean (M = 28.13) and standard deviation (SD = 8.99) and compare it to a normative (non-clinical) sample reported by Zimmerman et al. (2004) (M = 4.0; SD = 5.8). Based on this comparison, the clinical cut-off score was calculated to be 13.46. To identify those that did make clinical symptom change, we required participants to meet the following criteria: (1) a baseline score above the clinical cut-off (13.46), (2) a post-treatment score below the cut-off, and (3) at least a 50% reduction in MADRS score. This composite outcome helps confirm that the change is not only clinically significant but also exceeds what could be expected from measurement error alone. Dichotomous variables were coded for those making statistically significant symptom change (coded as 1) and those who did not (coded as 0) per the RCI. Next, Chi-square tests were used to compare rates of clinical and statistical improvement between treatment groups (ECT vs. ketamine). Additionally, a 2 (treatment type) x 2 (time: pre, post) mixed-ANOVA was conducted to assess changes in MADRS scores over time. All analyses were conducted in IBM SPSS Version 28.

3. Results

Results revealed that for the MADRS, 66 (45.2%) participants with multiple time points met the criteria for clinical symptom change and 80 (54.8%) did not. The results of the RCI revealed that over the course of treatment, 86 (58.9%) participants met the criteria for significant symptom change on the MADRS and 60 (41.1%) did not.

Next, results of Chi Square test revealed a significant association between treatment type and statistically significant change (RCI) on the MADRS [χ²(1) = 5.01; p = .025]. More specifically, patients who received ECT were more likely to have statistically significant change on the MADRS. Results of a Chi Square test examining the association between treatment type and clinical change on the MADRS [χ²(1) = 2.43; p = .119] were not significant.

A 2 (treatment type) x 2 (time) mixed model ANOVA was conducted to determine the effect of treatment type on MADRS score from the beginning to the end of treatment. Results revealed there was a significant interaction in treatment type over the course of treatment [F(1,133) = 9.19, p = .003 η² = .07]. More specifically, on the MADRS, results revealed that patients receiving ECT showed greater improvement on the MADRS, such that mean scores for the ECT group decreased from 26.87 (SD = 11.64) to 10.60 (SD = 8.78) from the baseline assessment to post-treatment assessment. The patients receiving ketamine showed improvement on the MADRS from a mean of 22.71 (SD = 9.22) to 12.57 (SD = 8.78) (see Figure 1 and Table 3). Response rates for ECT and ketamine were 67.3% and 45.7%, respectively, whereas remission rates for ECT and ketamine were 60.0% and 46.1%, respectively.

Figure 1. Average MADRS scores of ECT and ketamine from baseline (T1) to post-treatment (T2).

Error bars indicate SEM.

Table 3.

MADRS Scores by Treatment Type

	ECT		Ketamine
	M	SD	M	SD	Cohen’s d
MADRS T1	26.87a	11.64	22.71b	9.22	.59
MADRS T2	10.60a	8.78	12.57a	8.78	.22
Cohen’s d	1.77		1.13

MADRS: ECT (T1 n = 55; T2 n = 50), Ketamine (T1 n = 105; T2 n = 89). Means with the same superscripts within a row are not significantly different at p < .05. Means with different superscripts within a row are significantly different at the p < .05 level, as determined by post hoc tests. T1 = Baseline; T2 = post-treatment. Cohen’s d effect sizes are provided between ECT and Ketamine at both T1 and T2 as well as effect sizes to show the change from T1 and T2 within ECT and Ketamine.

4. Discussion

This retrospective chart review of open-label, nonrandomized treatment evaluated whether ECT or ketamine is efficacious for patients with MDEs. Our results showed significant differences between the effectiveness of ECT and ketamine over a three-week period. While both treatments led to reductions in depressive symptoms, ECT was associated with greater symptom improvement. These findings provide meaningful insights into the comparative utility between ECT and ketamine, suggesting that ECT may be a more effective treatment for targeting depressive symptoms (Folkerts et al., 1997; Kellner et al., 2020; Park et al., 2021).

We used the MADRS as our primary measure to assess treatment outcomes. Results revealed a significant association between treatment type and statistically significant change (i.e., symptom reduction), with patients who receeived ECT more likely to have significant symptom change compared to those patients who received ketamine. Interestingly, the association between treatment type and clinical symptom change was not significant, suggesting that while ECT may be more likely to lead to statistically significant symptom improvements, the difference in symptom improvement may not differ substantially between the two treatments. Nearly half of our overall sample demonstrated symptom reduction, reinforcing that while both interventions are effective, ECT may offer greater benefit.

Although we emphasized the antidepressant effects of ECT and using a large sample size, it is important to recognize the limitations of our results. First, our sample was relatively homogenous, with the majority of participants identifying as White ( 85.6% of the ECT group and 95.7% of the ketamine group) and female (60.3% of the ECT group and 61.7% of the ketamine group), which limits the generalizability of our findings to other diverse populations. Additionally, potential select bias should be considered. Our analysis did not control for setting-based differences that may be associated with variations in symptom severity, access to care, or insurance coverage. Second, due to the available data and our study design, we did not have an equal number of patients receiving each treatment, resulting in a higher number of patients receiving ketamine treatment than ECT. While our study provided demographic (sex, age, race), psychiatric history, and medication regimes at the time of treatment, we were unable to assess potential moderators of treatment response and tolerability (e.g., history of suicidal ideation, age of illness onset, inpatient vs. outpatient setting). Additionally, while medication regimens were collected at the start of treatment, any medication changes during the course of treatment were not controlled for in our analysis. These omissions may potentially limit our clinical characterization. Third, our data pool had missing outcomes, which could have potentially influenced the response rates reported in this dataset. Furthermore, we were not able to include the effects of any treatments beyond three weeks or follow-up outcomes assessed. Finally, patients were carefully monitored by psychiatrists and nursing staff during and after the ketamine or ECT treatment; however, our data did not include side effect information, including the impact on cognitive function. It is important to note that the patients in this sample were patients receiving clinical care and were not participants in a research study. As such, participants were not randomized to a treatment arm. However, the nature of the sample increased the generalizability of the results.

Our study supports the notion that ECT is a more effective treatment for depressive symptoms compared to ketamine. Our results also highlight the potential utility of ECT, not only for individuals with limited response to prior treatments but also as an efficient neuromodulation option over a 3-week period for TRD, due to its robust and rapid antidepressant effects. Although ketamine was less effective overall, it reamins as an effective treatment for some patients and may represent a clinically meaningful alternative for those with contraindications to ECT, a preference for less invasive treatment modalities, or medical and logistical barriers. Our findings highlight the importance of individualized treatment approaches, as some patients may respond better to ketamine. Given its greater availability in outpatient settings and less invasive nature, ketamine is often a more practical option for individuals who decline or cannot undergo ECT. Additionally, patients referred for ECT often present with more severe, treatment-resistant, or medically complex depressive episodes, which may have contributed to the observed differences in treatment outcomes between ECT and ketamine.

Future studies should continue to compare the effectiveness of ketamine and ECT using larger, more diverse samples, and should aim to include patient characteristics, treatment setting, and access-related factors that may moderate clinical outcomes. Additionally, the effects of combination therapy (e.g., receiving ECT and ketamine) and predictors of response (e.g., demographics, psychiatric/medical history, etc.) should be examined to better inform personalized treatment interventions.

Table 2.

MADRS Scores

	ECT		Ketamine		χ2	p
	n	%	n	%
Clinical	28	19.2	38	26.0	2.43	.119
Statistical	37	25.3	49	33.6	5.01	.025
Combined	28	19.2	38	26.0	2.43	.119

Clinical = the number of participants who met the criteria for Clinical Change as described in the data analysis section; Statistical = the number of participants who met the criteria for statistical significance as described in the data analysis section; Combined = the number of participants who met the criteria for both clinical and statistical significance.

Highlights.

• ECT and ketamine both led to a reduction in self-reported depressive symptoms.

• ECT exhibited the more robust antidepressant effects compared to ketamine.

• Randomized comparative effectiveness trials with large sample sizes may be required for generalizing our findings.