Introduction
The routine use of muscle relaxants in electroconvulsive therapy (ECT) starting in the 1950s has transformed the procedure from one with great morbidity (up to 35% vertebral fracture rate) [
[1]
] to one with few serious physical side effects [- Newbury C.L.
- Etter L.E.
Clarification of the problem of vertebral fractures from convulsive therapy: I. Incidence.
AMA Arch Neurol Psychiatr. 1955; 74: 472-478https://doi.org/10.1001/archneurpsyc.1955.02330170006002
[2]
]. ECT anesthesia involves short-acting general anesthetics to produce unconsciousness during the period of muscle relaxation [- Luccarelli J.
- Henry M.E.
- McCoy T.H.
Quantification of fracture rate during electroconvulsive therapy (ECT) using state-mandated reporting data.
Brain Stimul Basic Transl Clin Res Neuromodulation. 2020; 13: 523-524https://doi.org/10.1016/j.brs.2019.12.007
[3]
], and optimal anesthetic selection is an area of active research. Beyond initial anesthetic choice, at times during the ECT course anesthesia may also be changed, in particular with the intention of modifying seizure duration in the case of excessively short or long seizures. Given this, data on a large cohort of patients who undergo a switch in anesthetic during the ECT course may provide normative data to guide anesthesiologists and psychiatrists in considering an anesthetic change. This study analyzes the effects of switching anesthetic on seizure duration during acute course ECT.Methods
This was a single-center retrospective cohort study of patients aged 18+ receiving a first lifetime course of acute ECT from January 2000 to June 2017 and who had a change in anesthetic agent during the first twelve treatments. Treatments were performed using a Mecta Spectrum 5000Q (Tualatin, OR) with individualized seizure threshold determination [
[4]
] and subsequent suprathreshold treatments generally delivered thrice weekly [[5]
]. Dose and electrode placement were chosen clinically by the treating psychiatrist. Anesthetics used were methohexital, etomidate, propofol, and thiopental, with weight-based dosing and changes in anesthetic made at the discretion of the treating physicians. Succinylcholine was used as the muscle relaxant for all patients. Seizure duration was determined by the treating psychiatrist based on two lead bifrontal or fronto-mastoid EEG with propofol given at seizure termination or 120 seconds of continued seizure activity to reduce posttreatment agitation. For the primary analysis, the log-transformed seizure duration following anesthesia switch was modeled using linear regression, with prior seizure duration, sex, age (z-score), diagnosis (major depressive disorder, bipolar disorder, other), type of ECT [bilateral, right unilateral brief pulse (RUL-BP), right unilateral ultrabrief pulse (RUL-UBP)], applied ECT charge (z-score), muscle relaxant dose (z-score), anesthesia type prior to the switch (etomidate, thiopental, propofol, methohexital), and anesthesia type following the switch (etomidate, thiopental, propofol, methohexital) as predictors using R (v. 4, Vienna, Austria). Analysis was limited to anesthesia switches observed at least 10 times.Results
In total, 198 patients underwent anesthetic switches at mean of 6.9 ± 3.0 treatments (Table S1). Of these patients 86 (43.4%) were men, and mean age was 48.6 ± 15.4 years. Diagnoses were major depressive disorder (145; 73.2%), bipolar disorder (44; 22.2%), and other (9; 4.5%). At the treatment immediately preceding anesthetic switch 102 patients (51.5%) were receiving RUL-BP, 76 (38.4%) were receiving bilateral, and 20 (10.1%) were receiving RUL-UBP treatments. Following the anesthetic switch 86 (43.4%) patients received RUL-BP, 90 (45.9%) received bilateral, and 22 (11.1%) received RUL-UBP treatments. Anesthetics prior to switching were methohexital (116; 58.6%), propofol (46; 23.2%), thiopental (26; 13.1%), and etomidate (10, 5.1%), and following the switch were methohexital (49; 24.7%), propofol (24; 12.1%), thiopental (27; 13.6%), and etomidate (98; 49.5%).
The most common switches were methohexital to etomidate (n = 65), methohexital to thiopental (n = 27), and propofol to methohexital (n = 26). Analysis of the change in seizure duration for the treatments immediately prior to and following the anesthetic switch shows variable effects of the switches on seizure duration (Fig. S1). In a linear model of seizure duration following anesthesia switch, higher applied ECT charge was associated with shorter seizure duration, while age, sex, diagnosis, and electrode placement were not significantly associated (Table 1). Longer prior seizures were associated with longer seizures following anesthesia switch. Among anesthetic switches, transition from propofol was associated with an increase in seizure duration (estimate 0.21; 95% CI 0.01–9.42; p = 0.04), while transition to etomidate was likewise associated with longer seizures (estimate 0.24; 95% CI 0.05–0.44; p = 0.015). Other switches were not significantly associated with change in seizure duration.
Table 1Linear model of change in seizure duration before and after anesthetic switch.
| Predictors | Estimates | CI | p |
|---|---|---|---|
| Prior seizure duration (log) | 0.32 | 0.18–0.46 | <0.001 |
| Sex (male) | 0.03 | −0.11–0.18 | 0.646 |
| Age (z score) | 0 | −0.07–0.06 | 0.978 |
| Diagnosis | |||
| Major depressive disorder | −0.03 | −0.32–0.25 | 0.814 |
| Bipolar disorder | −0.12 | −0.42–0.18 | 0.447 |
| ECT Type | |||
| Right unilateral brief pulse | 0.15 | −0.04–0.35 | 0.127 |
| Bilateral | 0.09 | −0.12–0.29 | 0.407 |
| ECT charge (z score) | −0.07 | −0.13–−0.01 | 0.032 |
| Muscle relaxant dose (Z score) | −0.02 | −0.08–0.04 | 0.531 |
| Anesthetic prior to switch | |||
| etomidate | 0.06 | −0.28–0.39 | 0.731 |
| thiopental | 0.2 | −0.02–0.42 | 0.08 |
| propofol | 0.21 | 0.01–0.42 | 0.04 |
| Anesthetic following switch | |||
| etomidate | 0.24 | 0.05–0.44 | 0.015 |
| thiopental | 0.15 | −0.13–0.43 | 0.29 |
| propofol | −0.21 | −0.48–0.07 | 0.142 |
Bolded values are significant at the p=0.05 level.
Discussion
While there is extensive literature comparing ECT anesthetics [
[6]
,[7]
], the vast majority of these studies compare seizure durations among groups assigned to different anesthetics. In contrast, this study addresses switches in anesthetics for patients already undergoing ECT, a subject that has received minimal prior study. Our results compare treatments using methohexital, thiopental, etomidate, and propofol, yielding 12 potential pairwise switches of which eight occurred in 10 or more patients and are analyzed here. Both the transition from propofol and the transition to etomidate were associated with increase in seizure duration, while switches among other anesthetics were not associated with changes in seizure duration. These results largely match prior research on between-anesthetic differences, with meta-analysis indicating increased seizure duration using etomidate [- Fond G.
- Bennabi D.
- Haffen E.
- et al.
A Bayesian framework systematic review and meta-analysis of anesthetic agents effectiveness/tolerability profile in electroconvulsive therapy for major depression.
Sci Rep. 2016; 6https://doi.org/10.1038/srep19847
[6]
], and shorter seizures with propofol vs. methohexital [[7]
].- Fond G.
- Bennabi D.
- Haffen E.
- et al.
A Bayesian framework systematic review and meta-analysis of anesthetic agents effectiveness/tolerability profile in electroconvulsive therapy for major depression.
Sci Rep. 2016; 6https://doi.org/10.1038/srep19847
Overall these results suggest that with the exception of a switches in anesthetic from propofol or to etomidate, other switches are unlikely to cause significant changes in seizure duration. Our linear model further indicates that patient-level factors including age, sex, diagnosis, and ECT electrode placement do not significantly influence the change in seizure duration following anesthesia switch, while increasing the applied charge is correlated with shorter seizures as has been demonstrated previously [
[8]
].Strengths of this study include a large sample size. Additionally, consistency of ECT treatment team and protocols allows for clear comparisons among the patients receiving anesthetic switches, and reduces unmeasured sources of variability. The chief limitation is the retrospective cohort design, in which treatment decisions were made clinically rather than algorithmically. As a result, we are unable to assess why the treatment team decided to change anesthesia, and how the new anesthetic was chosen. As many anesthesia switches are likely made due to drug availability (for instance, many of the switches to thiopental in this study were as the result of a methohexital shortage at one point during the study period) or to avoid treatment adverse effects (e.g. cardiac events with thiopental) [
[9]
], this study is unable to examine the full spectrum of causes or effects of these anesthesia changes. Moreover we do not have standardized records of treatment efficacy or side effects, and are unable to quantify the effects of anesthesia switch on these outcomes.Conclusion
Based on 198 anesthetic switches during acute course ECT among methohexital, thiopental, etomidate, and propofol, the transition from propofol and the transition to etomidate were independently associated with lengthening of seizures.
Funding
This work was supported by the National Institute of Mental Health (R25MH094612, JL; R01MH120991, THM; 5R01MH112737-03, MEH) The sponsors had no role in study design, writing of the report, or data collection, analysis, or interpretation.
Declaration of competing interest
THM receives research funding from the Stanley Center at the Broad Institute, the Brain and Behavior Research Foundation, National Institute of Mental Health, National Human Genome Research Institute Home, and Telefonica Alfa. The remaining authors have no disclosures to report.
Appendix A. Supplementary data
The following is the supplementary data to this article:
- Multimedia component 1
References
- Clarification of the problem of vertebral fractures from convulsive therapy: I. Incidence.AMA Arch Neurol Psychiatr. 1955; 74: 472-478https://doi.org/10.1001/archneurpsyc.1955.02330170006002
- Quantification of fracture rate during electroconvulsive therapy (ECT) using state-mandated reporting data.Brain Stimul Basic Transl Clin Res Neuromodulation. 2020; 13: 523-524https://doi.org/10.1016/j.brs.2019.12.007
- Individualized anesthetic management for patients undergoing electroconvulsive therapy: a review of current practice.Anesth Analg. 2017; 124: 1943-1956https://doi.org/10.1213/ANE.0000000000001873
- Total charge required to induce a seizure in a retrospective cohort of patients undergoing dose titration of right unilateral ultrabrief pulse electroconvulsive therapy.J ECT. 2021; 37: 40-45https://doi.org/10.1097/YCT.0000000000000714
- Rate of continuing acute course treatment using right unilateral ultrabrief pulse electroconvulsive therapy at a large academic medical center.Eur Arch Psychiatr Clin Neurosci. 2021; 271: 191-197https://doi.org/10.1007/s00406-020-01202-2
- Evaluation of etomidate for seizure duration in electroconvulsive therapy: a systematic review and meta-analysis.J ECT. 2015; 31: 213-225https://doi.org/10.1097/YCT.0000000000000212
- A Bayesian framework systematic review and meta-analysis of anesthetic agents effectiveness/tolerability profile in electroconvulsive therapy for major depression.Sci Rep. 2016; 6https://doi.org/10.1038/srep19847
- Changes in seizure duration during acute course electroconvulsive therapy.Brain Stimulat. 2021; 14: 941-946https://doi.org/10.1016/j.brs.2021.05.016
- Induction of anesthesia with methohexital and thiopental in electroconvulsive therapy.N Engl J Med. 1965; 273: 353-360https://doi.org/10.1056/NEJM196508122730703
Article info
Publication history
Published online: July 19, 2021
Accepted:
July 16,
2021
Received in revised form:
July 6,
2021
Received:
June 8,
2021
Identification
Copyright
© 2021 The Author(s). Published by Elsevier Inc.
User license
Creative Commons Attribution – NonCommercial – NoDerivs (CC BY-NC-ND 4.0) | How you can reuse 
Elsevier's open access license policy
Creative Commons Attribution – NonCommercial – NoDerivs (CC BY-NC-ND 4.0)
Permitted
For non-commercial purposes:
- Read, print & download
- Redistribute or republish the final article
- Text & data mine
- Translate the article (private use only, not for distribution)
- Reuse portions or extracts from the article in other works
Not Permitted
- Sell or re-use for commercial purposes
- Distribute translations or adaptations of the article
Elsevier's open access license policy
