Stakeholders’ Ethical Concerns Regarding Psychiatric Electroceutical Interventions: Results from a US Nationwide Survey

Abstract

Background

Psychiatric electroceutical interventions (PEIs) use electrical or magnetic stimulation to treat mental disorders and may raise different ethical concerns than other therapies such as medications or talk therapy. Yet little is known about stakeholders’ perceptions of, and ethical concerns related to, these interventions. We aimed to better understand the ethical concerns of a variety of stakeholder groups (patients with depression, caregivers of patients, members of the public, and psychiatrists) regarding four PEIs: electroconvulsive therapy (ECT), repetitive transcranial magnetic stimulation (rTMS), deep brain stimulation (DBS), and adaptive brain implants (ABI).

Methods

We conducted a national survey of these four stakeholder groups, using an embedded video vignette experiment depicting a patient with treatment-resistant depression and her psychiatrist discussing the possibility of treatment with one of the four PEIs.

Results

Participants’ ethical concerns varied by stakeholder group, by PEI, and by the interaction of the two. The three non-clinician groups tended to have similar ethical concerns, but to differ from psychiatrists. Similar concerns were raised with regard to the two implantable technologies, DBS and ABI. Overall, there was relatively little concern about the involuntary use of PEIs, though some expressed concern about the adequacy of information provided during the consent process. There was also significant concern that patients may not receive helpful therapies.

Conclusions

To our knowledge, this is the first national survey that includes multiple stakeholder groups and multiple PEI modalities. A better understanding of the ethical concerns of stakeholders can help to shape clinical practice and health care policy regarding PEIs.

Keywords:

• Ethical concerns
• brain stimulation
• neuromodulation
• depression

Introduction

Psychiatric electroceutical interventions (PEIs) use electrical or magnetic stimulation to treat mental disorders. The neuroethics literature includes extensive discussion about the ethical issues raised by new PEIs that are largely experimental in psychiatric contexts, such as deep brain stimulation (DBS), and by already approved interventions, such as repetitive transcranial magnetic stimulation (rTMS). This literature has focused on various ethical issues raised by these therapies, including questions about their effectiveness and safety (Rossi et al. 2009; Greely et al. 2018; Cabrera et al. 2019), availability (Avalere Health and Mental Health America 2016; Cormier et al. 2019; Cabrera et al. 2022), informed consent (Blease 2013; Bell et al. 2014; Cabrera et al. 2022) and potential changes to personality and self (Johnstone 1999; Bluhm et al. 2021). However, there are significant gaps in our understanding of these ethical issues, for a number of reasons. First, there is relatively little research that connects the ethical issues raised by newer electroceutical interventions with longstanding concerns about older electroceutical therapies, particularly electroconvulsive therapy (ECT). Second, although there is some empirical work examining stakeholder perceptions of PEIs, it also focuses primarily on knowledge of and attitudes toward only one of these interventions, which limits the comparisons that can be made across different PEIs. Moreover, relatively few studies ask directly about participants’ views on ethical questions related to the use of PEIs.

One area of psychiatry where a better understanding of the ethical concerns related to PEIs is particularly important is the treatment of major depressive disorder (MDD). MDD is a serious mental health problem, with an estimated 7.2% of adults experiencing an episode of MDD in 2018 (SAMSHA 2019) and approximately a third of MDD patients not responding to treatment after courses of four different antidepressant medications (Warden et al. 2007). This high prevalence of treatment-resistant depression (TRD) suggests the need for other treatment approaches, of which PEIs may be a promising alternative. ECT and rTMS are currently FDA-approved treatments for MDD and are often used to treat TRD. ECT involves inducing seizure activity in the brain, while the patient is under anesthesia, through electrodes placed on the scalp. It is considered to be one of the most effective and safest therapies for TRD (Hermida et al. 2018; Rhee et al. 2021). A newer treatment, rTMS uses a coil placed on the head to alter neural activity; its effectiveness is also supported by the available evidence (Mutz et al. 2019). Both DBS and adaptive brain implants (ABIs, or closed-loop DBS systems) involve neurosurgery to implant electrodes in the brain. In DBS, stimulation levels are adjusted as needed by a clinician, while ABIs (which have just recently started to be investigated for TRD) are a variation on DBS that adjust the stimulation level in response to neural activity levels. Several randomized controlled trials have investigated DBS as a possible treatment for MDD, though its effectiveness is still unclear (Lozano et al. 2019).

As part of a larger project designed to investigate perceptions of PEIs among a variety of stakeholder groups, we surveyed four groups - patients with depression; caregivers for someone with depression; members of the public who did not fall into either of these categories; and psychiatrists - regarding their views on the ethical issues raised by four PEIs: ECT, rTMS, DBS, and ABIs.

Methods

Study design

Between April and June 2020, we gathered data from an online standardized survey with an embedded video vignette experiment. The survey was administered via Qualtrics to four large U.S. samples of depressed patients (n = 1050), caregivers for individuals with depression (n = 1026), members of the general public who were neither patients nor caregivers (n = 1022), and board-certified psychiatrists (n = 505). We used Qualtrics to draw our three non-clinician samples. The research team managed the recruitment and sampling of the psychiatrist stakeholder group. We used stratified random sampling, stratified by state, to select 16,190 U.S. psychiatrists listed in the American Board of Psychiatry and Neurology directory. Following the Tailored Design Method (Dillman, Smyth, and Christian 2014), we contacted this group multiple times via postal mail and email. Table 1 below displays key social and demographic characteristics for each of the four groups in this study.

Table 1. Description of the pooled sample (N = 3603).

	General Public (N = 1022)	Caregivers (N = 1026)	Depressed Patients (N = 1050)	Psychiatrists (N = 505)
Gender (% female)	50.9%	64.2%	73.7%	36.6%
Age (% aged 55+)	35.2%	21.9%	16.7%	60.4%
Race (% white)	78.0%	80.5%	88.7%	82.2%
Ethnicity (% Hispanic)	17.4%	13.2%	8.3%	5.7%
Education (% with bachelor’s degree)	45.2%	54.6%	33.3%	98.2%
Political Ideology (% liberal)	35.2%	40.0%	39.8%	63.6%
Religiosity (% monthly attendance)	35.0%	44.9%	30.9%	33.3%

Our survey was developed through a multi-stage process that incorporated insights from the relevant literature; pilot interviews with psychiatrists, patients, and members of the general public; four rounds of pilot testing; feedback from our Scientific Advisory Board composed of scholars with expertise in psychiatry, neuroethics, and ethics; and a round of cognitive testing with two patients and a psychiatrist. We employed a between-subjects 4*2 full factorial design (intervention vs severity of depression—moderate or severe) for our embedded experiment. Participants were randomly assigned to one of eight total conditions, and received the same set of core questions, with the addition of a few different questions unique to each stakeholder group. Michigan State University Institutional Review Board approved this study.

Procedures

After providing informed consent and answering several initial questions, participants viewed one of eight randomly assigned video vignettes, which featured professional actors and depicted a patient with either moderate or severe treatment-resistant depression (TRD) receiving information about a PEI (ECT, TMS, DBS or ABI) from her psychiatrist. Participants then answered questions assessing their understanding of the experimental message. They then answered a series of questions regarding their views about the PEI featured in the video they watched, as well as about their own demographic, social, and political characteristics.

Here, we present results from the portion of the survey asking participants to rank their top three (out of five items) ethical concerns related to their assigned PEI. The five items were: limited evidence of the treatment’s safety; treatment is too intrusive; patient may lack sufficient information for informed consent; treatment may be delivered without the patient’s consent; and patient not getting the treatment when it would actually help them. These five items were chosen based on salient themes from our developmental phase interview data and the neuroethics literature. Within the question, we randomized the item order to eliminate question order effects (Dillman, Smyth, and Christian 2014).

Analytical techniques

All analyses were conducted using IBM SPSS 26.0. We ran a series of multiple logistic regression models to examine variation in the top three ranked ethical concerns by stakeholder group, PEI modality (ECT vs. rTMS vs. DBS vs. ABIs), and depression severity (moderate vs. severe depression), while accounting for respondents’ demographic, social, and political characteristics. Since the distribution of respondents’ rankings violated the proportional odds assumption, using ordered logit regression was not justified. Instead, we employed logistic regression to predict whether or not respondents selected an ethical concern in their top three. We modeled the main effects of the experimental factors (i.e., PEI modality and depression severity) and stakeholder group membership (i.e., psychiatrists, patients, caregivers, and members of the public) with dummy variables, using ECT modality, moderate depression, and the psychiatrist group as the reference categories, respectively. To assess whether the influence of stakeholder group membership on ethical concerns was statistically moderated by PEI modality or depression severity, we included interaction terms in our multiple regression model.

Results

The odds ratios from the multiple logistic regression results in Table 2 provide evidence that participants’ ethical concerns vary by PEI modality (for all stakeholder groups) and across stakeholder group (for all PEIs). However, the differences across groups were largely driven by interactions between modality and stakeholder group, with the main differences between groups being found for the non-implantables (rTMS and, especially, ECT). We also observed different patterns when comparing ethical concerns between the implantable (DBS, ABIs) and non-implantable (ECT, rTMS) PEIs (Figure 1).

Figure 1. Panel showing the aggregate of the top three concerns across the four PEIs by stakeholder group.

Table 2. Odds ratios from multiple logistic regression models explaining top ethical concerns about use of PEIs across key stakeholder groups (N = 3603).

Predictors	Limited Evidence of PEI’s Safety	Treatment Is Too Intrusive	Patient May Lack Sufficient Information for Informed Consent	Treatment May Be Delivered without Patient’s Consent	Patient Not Getting Treatment When It Would Actually Help Them
Stakeholders
public (ref: psychiatrists)	5.43***	1.05	0.90	2.93***	0.39***
caregivers (ref: psychiatrists)	4.32***	1.16	0.82	4.16***	0.38***
patients (ref: psychiatrists)	4.05***	1.03	0.86	4.32***	0.39***
Experimental Conditions
rTMS (ref: ECT)	1.74***	0.75**	1.29*	0.67***	1.27*
DBS (ref: ECT)	2.20***	1.95***	0.83	0.69***	0.69***
ABI (ref: ECT)	2.64***	1.94***	0.85	0.62***	0.67***
severe TRD (ref: moderate)	0.93	0.89	0.85*	1.16*	1.15*
Perception of TRD
bad daily life with TRD	1.05*	0.99	1.01	.95***	0.99
Socio-Demographics
Female	1.32**	1.08	1.08	0.68***	0.91
Age	1.07*	0.96	1.04	0.91***	0.94**
White	1.24	0.97	0.97	0.86	1.04
educational attainment	0.96	0.99	1.00	0.92*	1.00
political ideology	1.03	1.01	1.02	0.99	0.99
Religiosity	0.94**	0.97	1.03	1.04*	1.05**
Nagelkerke R^2	.13	.06	.01	.10	.06

* p<.05.

** p<.01.

*** p<.001.

The supplemental file includes the disaggregation of the first, second and third choices, by PEI and stakeholder group.

PEI views by PEI modality and depression severity

Compared to participants assigned to the ECT condition, across all participant groups, those assigned to the any of the other PEI conditions were more likely to report limited evidence of the treatment’s safety among their top three ethical concerns, and less likely to report treatment may be delivered without patient’s consent. For the implantable PEIs, intrusiveness of the treatment was more often mentioned among the top three ethical concerns of participants than for ECT, patient not getting the treatment when it would actually help them was less often mentioned as a top ethical concern.

Depression severity had a very weak effect for ethical concerns, and only in three of the five possible items (see Table 2).

PEI views by stakeholder group

Compared with psychiatrists, participants in all three non-clinician groups were more than twice as likely to select limited evidence of the treatment’s safety and treatment may be delivered without the patient’s consent among their top three ethical concerns. They were also less likely to rank patient not getting the treatment when it would actually help them among their top three ethical concerns.

Influence of stakeholder group membership on PEI views by PEI modality

Table 3 reports the standardized coefficients from multiple OLS regression models. These are similar to those displayed in Table 2, but include a suite of interaction terms to investigate whether the influence of PEI modality or depression severity on ethical concerns varies across stakeholder groups. We draw attention to the most discernable patterns here.

Table 3. Odds ratios from multiple logistic regression models explaining top ethical concerns about use of PEIs across key stakeholder groups (N = 3603).

Predictors	Limited Evidence of PEI’s Safety	Treatment Is Too Intrusive	Patient May Lack Sufficient Information for Informed Consent	Treatment May Be Delivered without Patient’s Consent	Patient Not Getting Treatment When It Would Actually Help Them
Stakeholders
public (ref: psychiatrists)	5.21***	1.01	0.89	3.62***	0.31***
caregivers (ref: psychiatrists)	4.08***	1.11	0.82	5.13***	0.30***
patients (ref: psychiatrists)	3.87***	0.98	0.86	5.31***	0.31***
Experimental Conditions
rTMS (ref: ECT)	1.87***	0.75**	1.28*	0.63***	1.21
DBS (ref: ECT)	2.15***	2.01***	0.83	0.62***	0.62***
ABI (ref: ECT)	2.58***	1.96***	0.85	0.55***	0.60***
severe TRD (ref: moderate)	0.91	0.89	0.85*	1.13	1.20*
Perception of TRD
bad daily life with TRD	1.06**	0.99	1.01	0.95***	0.99
Socio-Demographics
female	1.28*	1.05	1.08	0.68***	0.94
age	1.08*	0.96	1.04	0.91***	0.94**
white	1.24	0.97	0.98	0.86	1.05
educational attainment	0.96	0.99	1.00	0.91*	1.00
political ideology	1.03	1.01	1.02	0.99	0.99
religiosity	0.93**	0.97	1.03	1.04*	1.06**
Interaction Terms
rTMS* public	0.13***	0.81	1.44	3.51**	2.34
rTMS* caregiver	0.07***	1.28	1.51	3.30**	1.55
rTMS* patient	0.15***	0.87	1.36	2.05	1.92
DBS* public	0.04***	0.32***	1.18	3.78**	7.95***
DBS* caregiver	0.02***	0.28***	1.37	4.47***	7.23***
DBS* patient	0.03***	0.21***	1.11	4.21***	7.67***
ABI* public	0.01***	0.47*	2.02*	5.11***	8.28***
ABI* caregiver	0.01***	0.68	1.24	4.61**	8.82***
ABI* patient	0.01***	0.49*	1.25	4.93***	10.54***
severe* public	0.95	1.24	1.13	1.46	0.51**
severe* caregiver	1.36	1.50	0.85	1.47	0.39***
severe* patient	1.85*	1.01	1.26	1.19	0.52**
Nagelkerke R^2	.21	.07	.02	.12	.08

* p<.05.

** p<.01.

*** p<.001.

First, assignment to the ECT condition moderates limited evidence of the treatment’s safety between psychiatrists and non-clinician groups. Specifically, in the ECT condition, non-clinician group participants in the ECT condition were more likely than psychiatrists to select limited evidence of the treatment’s safety among their top three ethical concerns. Non-clinician group participants in the rTMS condition were more likely than psychiatrists to select treatment may be delivered without the patient’s consent among their top three ethical concerns, whereas for the DBS or ABI condition, the non-clinician groups were all more likely to select treatment may be delivered without the patient’s consent and patient not getting the treatment when it would actually help them.

Finally, for the non-clinicians, limited evidence of the treatment’s safety was among the top three ethical concerns selected, for ECT and rTMS, whereas for psychiatrists it was patient not getting the treatment when it would actually help them. For the implantable PEIs, we saw a very similar pattern, with limited evidence of the treatment’s safety and treatment is too intrusive among the top three ethical concerns selected. The only item with a difference between clinicians and non-clinician groups was treatment may be delivered without the patient’s consent, but for both groups that was the item least likely to be selected among the top three ethical concerns. Overall, these results provide considerable evidence that ethical concerns differ strongly between psychiatrists and non-clinician groups for the already approved PEIs, particularly ECT, and among different implantable and non-implantable modalities of PEIs.

PEI views by socio-demographic characteristics

For the most part, sociodemographic variables accounted for little of the adjusted R² values across the models. Compared to their male counterparts, females were more likely to perceive limited evidence of the treatment’s safety among their top three ethical concerns, and less likely to perceive treatment may be delivered without the patient’s consent. Increased age is related to a lesser likelihood of ranking treatment may be delivered without the patient’s consent and patient not getting the treatment when it would actually help them among top ethical concerns. Finally, educational attainment, political ideology, and religiosity had only weak associations with just a few ethical concerns (see Table 2). We did not find an association between race and ethical concerns.

Discussion

Our results suggest that ethical concerns about PEIs are influenced primarily by the specific PEI modality, and by the interaction between participant group membership and PEI modality. Here, we discuss the effects of these variables on the selection of different ethical concerns.

Safety

Safety is an important ethical issue, which has been raised as a concern by the clinical rTMS community (Rossi et al. 2009), patients regarding rTMS (Mayer et al. 2012), patients receiving DBS for TRD (Lawrence et al. 2018), and also in online comments regarding PEI use (Cabrera et al. 2019). Our results show a striking difference in perception of safety between psychiatrists and non-clinicians regarding the two approved modalities, particularly ECT. Non-clinician stakeholder groups selected limited evidence of the treatment’s safety as their top ethical concern for both ECT and rTMS, while psychiatrists did not prioritize this concern for either of the approved modalities. This gap in belief regarding the safety of ECT persists despite strong evidence of treatment safety (Tørring et al. 2017). Previous work has suggested that negative portrayals of ECT in movies and the media more broadly may influence how it is perceived by individuals with no personal experience of the treatment (Griffiths and O’Neill-Kerr 2019; Cabrera et al. 2021c). Our findings are also consistent with previous studies that found that favorable views of ECT increase with more education and exposure to the treatment, and are greater in medical students compared to the public (McFarquhar and Thompson 2008), and in psychiatrists compared to other mental health professionals (Lutchman et al. 2001).

Similar to ECT, the non-clinician group participants in the rTMS condition ranked safety as a top concern. These safety concerns are perhaps surprising given other results that have shown a patient preference for rTMS over ECT (Magnezi et al. 2016). While psychiatrists did not see safety as a main concern for TMS, there were important differences with their safety concerns around ECT. These results suggest that some psychiatrists believe the data for safety in rTMS is lacking compared to that of ECT. This may be due to the relatively short history of rTMS as a treatment for MDD, and, relatedly, among older psychiatrists, to the lack of exposure to this therapy during training (Stern et al. 2016; Cabrera et al. 2021b). Further, it is interesting to note this difference in concerns about safety considering different risks associated with each treatment modality. ECT, but not rTMS, requires general anesthesia which carries some risk associated with using anesthesia as well as ECT’s potential negative effects on memory and cognition (Robertson and Pryor 2006); however, rTMS carries a small risk of an induced seizure (Dobek et al. 2015).

For the implantable modalities, all stakeholders ranked limited evidence of the treatment’s safety high for both DBS and ABIs. This is likely because psychiatrists recognize that they are still in the experimental stage and are less familiar with these unapproved modalities. Another possible explanation for our findings is that, in general, people view implantable PEIs as inherently riskier (Wilhelmy et al. 2018) and as having more side effects than other treatments for depression, particularly antidepressant medications (Cabrera et al. 2021a).

Patients not receiving the treatment when it would help them

Whereas the top ethical concern related to the two approved PEIs for all non-clinician groups was limited evidence of the treatment’s safety, among psychiatrists, it was patient not getting the treatment when it would actually help them, with a majority of psychiatrists selecting this option as their first-ranked concern for both ECT and rTMS. There are several reasons that this might be the case. First, there may be structural barriers in place that prevent patients from accessing these therapies; these include geographic limitations (Dauenhauer, Chauhan, and Cohen 2011; Cabrera et al. 2022), problems with insurance coverage (particularly for rTMS) (Rasmussen 2011; Cabrera et al. 2022), and the requirement by many insurance companies that do cover these treatments that multiple trials of medication therapy must be attempted before treatment with a PEI is covered (Hoffman 2018). Second, even among psychiatrists, there appears to be disagreement about whether ECT or rTMS should be available earlier or later in the sequence of treatments. Cabrera et al. (2021a, 2021b), reported that while most of the psychiatrists they interviewed agreed that PEIs should not be first-line treatments for MDD, there was disagreement about how many medication trials should be attempted before offering ECT as an alternative.

With regard to the two implantable PEIs, psychiatrists’ ranking of patient not getting the treatment when it would actually help them as an ethical issue was more in line with the ranking of the non-clinician groups. The difference in response for the approved (nonimplantable) PEIs and the experimental (implantable) PEIs likely reflects the established efficacy of ECT and rTMS relative to that of DBS and ABIs, as well as the view that these riskier interventions should be left as last resort treatments. It should also be noted, however, that for all groups and PEIs (other than psychiatrists’ high rankings of this item for ECT and rTMS), between 40 and 60% of participants, in all groups, ranked this issue in their top three concerns. This result may reflect a general sense among our participants that effective therapies should be made available to patients, rather than something specific to PEIs.

Informed consent and involuntary treatment

Two of the items in our survey were related to consent. One item emphasized the importance of providing potential patients with sufficient information to give valid consent, while the other asked about the potential for the use of the PEI in involuntary treatment.

Patient may lack sufficient information for informed consent was a concern for all modalities across all stakeholder groups. Our findings echo the neuroethics literature on informed consent as a concern for patients and providers regarding approved PEIs, such as ECT (Blease 2013) and rTMS (Rossi et al. 2009). Further, Cabrera et al. (2022) found that the most common theme raised in response to a question about the ethical issues raised by PEIs was informed consent, cutting across all stakeholder groups (Cabrera et al. 2022). These interview data also shed further light on the complexities of this concern, including issues about lack of information available (echoing concerns about lack of evidence about safety) and the actual provision of information in ways that empower patients to make a choice that reflects their own needs and values, such as full disclosure and discussion of uncertainties with regard to the therapies under discussion (Park et al. 2017). This information provision is especially important because assessments of risk inherently contain a value component in addition to consideration of the probability of different outcomes (Szmukler 2014).

In the case of the implantable PEIs, concerns about informed consent are also raised in the neuroethics literature. For example, it is a guiding neuroethics principle for the NIH Brain initiative, specifically to “anticipate special issues related to capacity, autonomy, and agency” (Greely et al. 2018), including the ability of potential patients (or, currently, research participants) to consent to treatment with technologies like DBS and ABIs. Although concerns have been raised that severe, treatment-resistant depression might diminish the capacity to weigh information relevant to consent, it may not be easy to determine whether this is the case. For example, Lawrence et al. (2018) found, in interviews asking participants what would influence their decision about whether to participate in a DBS trial, that inpatients with depression were able to weigh the risks and potential benefits of trial participation. At the same time, however, some of their participants doubted their own ability to do so. Specific concerns regarding these implantable interventions have also been elicited among researchers that include these being viewed as a last resort, or that patients may feel pressure to participate in clinical trials (Muñoz et al. 2020). Additionally, the continuous and autonomous nature of ABIs raise concerns of informed consent.

Our data show that although informed consent is an important concern for all modalities; the concerns raised specific to implantable devices did not translate to more concern for adequate informed consent compared to the approved modalities. This may, at least for the non-clinician groups, be because our vignettes did not present the two implantable PEIs as experimental therapies.

Although all the stakeholder groups were concerned that patients would not have sufficient information to give valid consent to treatment with a PEI, this issue is distinct from the question of involuntary treatment (i.e., treatment against the patient’s wishes). None of the stakeholder groups rated the possibility that treatment may be delivered without the patient’s consent as a top ethical issue; however, direct comparison of stakeholder groups showed that non-clinician groups were more likely to do so than were psychiatrists. Our results are surprising, given longstanding concerns about involuntary treatment in psychiatry.

The literature on involuntary treatment is mostly focused on ECT, due to its historical involuntary use and subsequent media portrayals. A survey by Teh, Helmes, and Drake (2007) showed strong opposition among the Australian public to using ECT for involuntary patients. Survey respondents commonly cited freedom of choice and the importance of informed consent as concerns. It is of note, however, that the vignette used by Teh et al. portrayed an individual refusing ECT, which may have primed participants to focus on the possibility of involuntary use of the treatment. By contrast, our vignette depicted a conversation between a patient and her psychiatrist in which the PEI was presented as a possible next step in therapy and the patient was clearly willing to discuss the treatment further.

Reluctance to administer involuntary ECT has also been shown among psychiatrists. In a study by Besse et al. (2019), psychiatrists in Germany reported that they would be less likely to consider ECT in involuntary patients, compared with other measures, such as antipsychotic and benzodiazepine pharmacotherapy. Their participants cited concerns that included fear of being sued and invasiveness of the treatment. The psychiatrists in our study may also favor medication approaches in these circumstances, which would explain why they did not perceive involuntary treatment with ECT among their top three ethical concerns.

Overall, our findings with regard to the two consent items must be interpreted in the context within which ECT is actually used. More than any other treatment for MDD, the use of ECT is restricted by legal requirements, which are set in the U.S. at the state level. For patients with decision-making capacity, these laws may influence the informed consent process by requiring specific information to be provided, notably, in some states, the fact that ECT is controversial (Livingston et al. 2018). Yet given the negative public perception of ECT, this requirement may lead to patients weighing the risks of ECT much more heavily than the well-established benefits of the therapy. For patients who lack decision-making capacity, laws may require additional court authorization, in addition to the consent of the patient’s proxy decision-maker (Livingston et al. 2018; Iltis et al. 2023). As Iltis et al. note, these requirements may delay, or even prohibit, patients’ access to a needed therapy, and may therefore be detrimental to their mental health.

Intrusiveness

Unsurprisingly, across all groups, treatment is too intrusive was more likely to be ranked as an ethical concern for the implantable versus the non-implantable technologies. This is most likely because both DBS and ABIs are indeed invasive treatments, as they require neurosurgery and the implantation of electrodes that remain in the brain. This understanding of the concept of “invasiveness” reflects the common distinction in the neuroethics literature between invasive and noninvasive interventions (Emerging Issues Task Force and International Neuroethics Society 2019; Hendriks et al. 2019), with the later generally considered to be riskier. However, there is also some suggestion that the concept is more complex than this simple distinction would suggest. For example, Rudnick (2011) has argued that it is best understood in terms of the ability of an intervention to cause harm. Bluhm et al. (2023) found that participants in an interview study considered a number of factors, including the permanence of an intervention, how familiar it seemed, and the amount of control a patient retained over the intervention and its effects, in making judgments about how invasive they believed a PEI to be. In addition, they considered the invasive effects of the PEI on a person’s emotional state and lifestyle, in addition to its (relatively) straightforward physical invasiveness.

Overall, our results may be informative for the development of policies and guidelines for the provision of PEI therapy – both the currently available treatments and the implantable modalities if they are approved. First (bearing in mind the legal requirements, in the case of ECT), knowledge of the ethical concerns of patients and potential users of PEI treatments should inform guidance for discussions of informed consent between patients and providers. If and when implantable PEIs are approved, similar care should be taken to ensure that patients are given clear and accurate information about their benefits and risks. Second, our results build on previous research suggesting that there is disagreement among clinicians regarding when these therapies should be offered. Any new treatment guidelines for MDD should explicitly address how PEIs fit into a treatment plan, including patient-specific factors that might influence when they should be offered.

This study has several limitations. First, only four PEIs were included in the study, which were selected based on characteristics such as their technological features, risk-benefit profile, and historical background. Our findings cannot be generalized to other interventions, such as vagus nerve stimulation and transcranial direct current stimulation, which may raise different ethical concerns. Similarly, our results focus solely on MDD and do not necessarily generalize to other psychiatric disorders for which PEIs might be used. Finally, we asked participants to select their top three ethical concerns from a list of five ethical issues that we had identified as particularly important based on our earlier interviews and the relevant neuroethics literature. This methodological choice facilitated comparison across participant groups and PEIs, though it did so at the expense of the comprehensiveness that could be achieved with a free response option that could not be readily compared.

In conclusion, our results suggest that the ethical concerns raised by PEIs vary with the specific PEI in question, with stakeholder group, and with their interaction. The greatest differences were observed between our psychiatrist group and the three non-clinician groups, which tended to be more similar. Depression severity did not exert a large influence on participants’ ethical concerns.

Ethics approval information

This study was approved as exempt by the Michigan State University IRB (STUDY00001247).

Acknowledgements

We thank Patricia Henegan for her help with proof reading.

Conflict of interests

LC, ES, RB, AM have no conflicts of interest to declare. Dr. Achtyes has served on advisory boards for Alkermes, Janssen, Lundbeck/Otsuka, Roche, Sunovion, Teva and CAPNOSZero (unpaid), reports previous stock holdings in AstraZeneca, Johnson & Johnson, Moderna, and Pfizer. Dr. Achtyes has received research support from Alkermes, Astellas, Biogen, Boehringer-Ingelheim, InnateVR, Janssen, National Network of Depression Centers, Neurocrine Biosciences, Novartis, Otsuka, Pear Therapeutics, Takeda and serves on the SMI Adviser LAI Center of Excellence (unpaid). This study was supported by a BRAIN Initiative grant (#RF1MH117802) from the US National Institute of Mental Health.

Additional information

Funding

This study was supported by a BRAIN Initiative grant (#RF1MH117802) from the US National Institute of Mental Health to (PI: LC). The funding source had no influence on the study design; the collection, analysis, and interpretation of data; the writing of this manuscript; or the decision to submit the manuscript for publication.