Jumping through the hoops: Barriers and other ethical concerns regarding the use of psychiatric electroceutical interventions

1.1. Structural barriers

Structural barriers are bureaucratic, economic, or infrastructural factors associated with health services, including availability of services and providers, administrative challenges, limited resources, cost of therapy, and referral patterns by psychiatrists and other clinicians that affect patients’ access to a therapy (Tomczyk et al., 2020). A key structural barrier to medical treatments in general, and mental health care in particular, is access to therapies (HealthMental Health America, 2016). Financial costs and insufficient insurance coverage are other dimensions of access barriers in mental health care (Rowan et al., 2013). Access issues are exacerbated by inconvenience, or inability of patients to obtain an appointment (Mojtabai et al., 2010), transportation limitations, space limitations reducing the numbers of people treated (Mechanic, 2002, Wilkinson et al., 2021), as well as the necessity to be released from work for treatment, and arranging childcare.

A barrier specific to PEIs is the fact that medical education does not include sufficient focus on PEIs (Goldbloom and Gratzer, 2019). The lack of clarity in PEI clinical guidelines (APA, 2001; McClintock et al., 2018; Perera et al. 2016), including who is an optimal patient and when to start PEI treatment, is another structural barrier (Cormier et al., 2019). Related to this is the infrequent use of clinical practice guidelines by providers for clinical decision-making (Cabana et al., 1999; HealthMental Health America, 2016), which can also deter the expedient use of evidence-based practices. Moreover, in the specific case of PEIs, current guidelines for ECT (APA, 2001) and TMS (McClintock et al., 2018; Perera et al. 2016) are outdated.

A final major structural barrier is the lack of a federal legal framework providing guidance on PEI procedures, including who can receive these interventions and who is qualified to prescribe or perform the procedures (Livingston et al., 2017).

1.2. Attitudinal barriers

Attitudinal barriers include beliefs, attitudes, and emotions that inhibit help-seeking and treatment (Tomczyk et al., 2020). These include stigma (Mojtabai et al., 2010; Mental Health America, 2016), negative beliefs or fear of treatment (e.g., concerns about memory loss), and misinterpretations or uncertainty about treatment consequences (Andrade et al., 2013).

For example, while stigma is a common barrier to the use and public acceptance of mental health treatments, this is exacerbated in the case of some PEIs, in particular given the negative image of ECT among the general public (Lauber et al., 2005; Wilkinson et al., 2021). In the case of PEIs involving neurosurgery, such as DBS, views towards them might be influenced by a history of transgressions and complications related to the use of older ablative procedures performed in the mid-1900s. Resistance from psychiatrists and other clinicians to referring patients is another attitudinal barrier (Tomczyk et al., 2020). As such, there are mixed perceptions among both clinicians and the general public (Cormier et al., 2019; Cabrera et al., 2018a).

1.3. Other ethical issues

Beyond structural and attitudinal barriers, the neuroethics literature has identified other important ethical issues when discussing PEIs. Many of the ethical concerns identified in the neuroethical literature for PEIs are also relevant for medications as they are for PEIs, including concerns about their effectiveness (Cabrera et al., 2019; 2018b), the balance between benefits and risks, informed consent (Bell et al., 2014; Fins 2009; Blease 2013; Rossi et al., 2009), and effects of these interventions on personal identity (Baylis, 2013; Jotterand and Giordano, 2011; Lipsman and Glannon, 2013), personality (de Haan et al., 2017; Synofzik and Schlaepfer, 2008), agency (Glannon, 2014; Goering, 2014), autonomy (Goering, 2015, 2017), and authenticity (Johansson et al., 2011; Kraemer, 2011). However, the similar list of issues raised for PEIs and for pharmaceutical treatments does not mean that stakeholders perceive a specific issue to be equally important for both kinds of therapies. Moreover, the fact that PEIs are considered for treatment resistant depression (TRD)—a form of depression that does not improve despite the use of multiple and appropriately dosed antidepressant medications, and that we know less about their mechanisms of action, do make some of these issues more salient or bring different considerations into play.

A better understanding of the barriers to, and other ethical concerns about, PEIs is important for planning mental health services, setting resource allocation priorities, reducing the burden of mental illness (Andrade et al., 2013) and addressing misconceptions. However, little research has examined stakeholders’ views about barriers and other ethical considerations around these interventions, and the few that exist are mostly on ECT and are focused only on one stakeholder group. This study aims to provide insights into this important question with a focus on electroceuticals for TRD among three different stakeholder groups—psychiatrists, persons diagnosed with depression, and members of the general public.

2.1. Recruitment

We recruited practicing psychiatrists who had at least some experience with and/or interest in neuromodulation techniques. We emailed potential participants recommended by a member of the research team with an invitation to participate in the study. If we received no response within two weeks, we sent a reminder email, after which we made no further contact. When selecting psychiatrists to interview, we aimed for a sample of at least 20% females, at least one non-white participant, and a range of clinical experience with PEIs, with at least one psychiatrist with substantial or moderate experience with either ECT, TMS, or DBS. We recruited potential patients diagnosed with depression via flyers in mental health clinics primarily, but we also advertised via local establishments (e.g., coffeeshops and libraries), and social media posts. We recruited members of the public via local establishments (e.g., coffeeshops and libraries), and social media posts. We pre-screened participants, asking them if they had ever been diagnosed with depression, to ensure we included them in the correct group.We followed up by email or phone with those individuals who indicated interest in the study. For our patient and public samples, we aimed for a balance of males and females, and race and ethnicity, and we aimed to have at least five patients with moderate experience with a PEI, and five members of the public with caregiving experience. Table 1 provides selected demographic characteristics of the three groups. For details on participant’s experience or knowledge level with PEIs please see the supplemental material.

2.2. Procedures

When scheduling interviews, we sent participants a consent letter containing key details about the study and a brief fact sheet summarizing the PEIs (see supporting information S1 document). Prior to each interview, we invited participants to ask any questions they had about our study before securing their oral consent to participate.

We created our interview guides using insights from relevant prior studies and feedback from our project’s Scientific Advisory Board. The interviews took place over a period of four months from February to May 2019. Semi-structured interviews took place in person, via Zoom, or by telephone. We started the interviews with general questions about the participants’ demographic information, background, and experience. We also asked patients and members of the public to describe the PEIs to us to help ensure that they understood the therapies being discussed. The remainder of the interview guides share the same core questions; each participant was asked about either one or two PEIs (depending on time limits), selected based on the participant’s experience and interest, while also keeping a balance across the PEIs in the study. The interviews ranged from 30 to more than 60 minutes. Participants were compensated for their time completing the interviews.

While these interviews covered a range of PEI topics, we focus this paper on perceptions of barriers to and other ethical concerns about using PEIs (See supporting information S2 Document). All interviews were recorded, anonymized, and transcribed verbatim for analysis.

2.3. Data analysis

We analyzed our data employing a qualitative content analysis approach (Hsieh and Shannon, 2005; Sandelowski, 2000), using the web-based software Dedoose. We developed a codebook with a priori codes based on key insights from the relevant neuroethics literature, then added new codes emerging from a few interview transcripts. We identified any necessary adjustments to the codebook, using the method of ‘constant comparison’ to structure the coding framework—comparing emerging codes with established codes merging similar ones together. Finally, we developed broader categories with codes being merged together to begin the process of identifying themes or emerging concepts.

Two team members (MC and EC) coded the remaining interviews with the finalized codebook, and another two (LC and RB) served as secondary coders. We used ‘interpretive convergence’ for code validation, with intensive group discussion of codes until consensus was achieved (Saldana, 2008) in cases of disagreement. We kept a research log during data analysis to enhance rigor and encourage transparency of research decisions within the process and reflection during analysis.

While barriers are an ethical issue, we decided to separate other ethical issues from structural and attitudinal barriers (Table 2).

We include selected interview excerpts to illustrate key themes and sub-themes. For readability, we have removed any non-content utterances.

3.1. Barriers

The most common barrier across all groups was access limitations, followed by perceived low mental health literacy. Time constraints was also a commonly mentioned barrier, but the presence of this theme varied considerably across the stakeholder groups, with psychiatrists more frequently reporting this as a barrier than members of the public. Similarly, compared to other stakeholder groups, psychiatrists more frequently identified concerns about the limitations for caregiver support.

3.2. Other ethical concerns

The most common theme for other ethical concerns within the entire sample was a concern about informed consent. The second most common theme was perceived lack of safety from side effects and unanticipated long-term risks. Limited efficacy of PEIs was also commonly mentioned, but primarily by psychiatrists compared to other groups. The final theme that emerged was a broader concern about social justice.

4.1. Limited access

Our results suggest that electroceuticals may face stronger structural barriers to use than other types of mental health treatments, though further research is needed to confirm this claim. Our participants believed that PEIs are even less likely to be available than other mental health treatments, due to the lack of equipment and of providers who offer these interventions, including geographical proximity to clinics that provide PEIs.

In the case of implantable PEIs, which will require even more specialized neurosurgical treatment and equipment, geographical distance to medical academic centers where these procedures are likely to be offered is a further limitation (Goldbloom and Gratzer, 2019). Additionally, not all psychiatrists have been trained sufficiently on PEI procedures. In fact, the Accreditation Council for Graduate Medical Education only requires that psychiatry residents demonstrate knowledge of ‘indications for’ and ‘uses of’ electroconvulsive and neuromodulation therapies as a graduation requirement (ACGME, 2019).

Moreover, the US has a noticeable shortage of psychiatrists (National Council Medical Director Institute, 2017), and an increasing demand for mental health treatments. As previous studies have found (Tomczyk et al., 2020), a lack of provider referrals to PEI interventions likely contributed to the limited availability of PEIs, especially regarding surgical psychiatric interventions, such as DBS (Cormier et al., 2019; M. I. Hariz and G. M. Hariz, 2012). Since TMS was FDA-approved for depression in 2008, the number of private TMS clinics has increased (Hutton, 2014). However, this increase may not address the access problem for everyone, since these private clinics tend to be located in densely populated areas rather than in rural areas that are medically underserved. Private clinics can also lead to increased variability in administrative processes, treatment guidelines, insurance coverage, and cost.

Financial barriers to mental health treatment have increased historically (Mojtabai et al., 2010, Rowan et al., 2013) and were mentioned by all three stakeholder groups. Psychiatrists were concerned with barriers related to insurance coverage and prior authorizations, while patients paid the most attention to actual out-of-pocket costs. For example, while TMS has substantially lower out-of-pocket costs than other PEIs (without insurance coverage, the out-of-pocket cost of a course of TMS is approximately $10,000 (Rasmussen, 2011)), insurance companies are still reluctant to cover the procedure. Newer TMS devices and protocols are being established which might not be FDA-approved, further complicating reimbursement. Despite the increased barrier for insurance coverage of TMS compared to ECT, some studies suggest that TMS is more cost-effective than ECT (Fitzgibbon et al., 2020) and than pharmaceuticals (Nguyen and Gordon, 2015) for TRD.

Additionally, several insurance companies require a stepped protocol, in which patients must “fail” several lower-cost treatments, such as medication, before trialing more expensive options (Hoffman, 2018), which could further discourage the use of PEIs. The 2008 Mental Health Parity and Addiction Equity Act (https://www.congress.gov/bill/110th-congress/house-bill/1424) requires insurance coverage for mental health conditions to be no more restrictive than coverage for other medical conditions (Busch, 2012). However, even today, many insurance plans in the US have failed to comply with all aspects of this law (Davenport et al. 2019), particularly with regard to prior authorizations, reimbursement rates, and network designs. While several advocacy and medical groups have supported the enforcement of the law (e.g., American Psychiatric Association), fair and equitable insurance coverage still remains an issue for those seeking treatment for mental health disorders, further reducing access to these types of treatments.

4.2. Low mental health literacy

Mental health literacy is associated with individuals’ decision-making regarding mental health, especially help-seeking behaviors for psychiatric symptoms (Kutcher et al., 2016; Furnham and Swami, 2018).

Despite much effort by health organizations to raise mental health literacy, the public is still only vaguely aware of most types of psychiatric disorders and the availability of effective treatments. Similar to our study, Cormier and colleagues found that approximately 83% of their sampled psychiatrists and residents considered their level of knowledge of DBS to be weak or poor (Cormier et al., 2019). It is possible that DBS may be particularly prone to knowledge deficits, given its investigational status.

Stigma associated with seeking specialty mental health services prevents many individuals living with depression or other severe mental illnesses from pursuing and obtaining adequate psychiatric care (Mechanic, 2002). Consistent with prior studies (McFarquhar and Thompson, 2008; Wilhelmy et al., 2018, Wilkinson et al., 2021), our stakeholder groups perceived PEIs, especially ECT, to be especially stigmatized. Media-driven stereotypes of psychiatric treatments—especially ECT—endure, likely reducing public recognition of improvements in how procedures are performed.

4.3. Time constraints and limitations on caregiver support

To the best of our knowledge, no studies have substantially investigated time constraints as a barrier to PEI treatment, though Kessler et al. (2001) found that almost one-third of patients did not seek mental health treatment in general because it takes too much time.

Approved TMS protocols typically involve four to five sessions per week for four to six weeks, equalling 20–30, 45-minute treatment sessions (Rasmussen, 2011), and our participants frequently identified time commitment as a potential barrier. To receive treatment so frequently requires one to have adequate and reliable transportation, funds for travel, time available from work or other life responsibilities. While newer TMS modalities, such as low-frequency magnetic stimulation, are less time-consuming (Somani and Kar, 2019), they also tend to be less available and not covered by insurance.

Even though ECT is administered less often (usually six to twelve sessions conducted two or three times weekly (American Psychiatric Association, 2010), the time required for anesthesia and recovery is nevertheless a substantial barrier. Also the general anesthesia used during ECT requires a patient to have someone transport them to and from the treatment facility. While our participants did not mention these themes regarding DBS, the fact that this intervention involves a neurosurgical procedure creates the need for a caregiver not only for transportation, but also potentially for physical and emotional support during recovery. Moreover, there are pre- and post-operative follow-up visits, including several evaluations, and consultations, for battery replacement (Ramirez-Zamora et al., 2018), which are time-intensive.

4.4. Neuroethical considerations

Informed consent is both an ethical and a legal obligation that allows patients to make educated decisions about their own health care. Its legitimacy requires that patients fully understand the information available to inform their decision-making. How PEIs work when treating depression, their effectiveness, and their long-term side effects are not fully understood. For example, if there is limited evidence of effectiveness, we may be putting patients at risk of not getting appropriate treatment, and they may experience side effects without compensating treatment benefits. This underscores the importance of ensuring that healthcare professionals are transparent when discussing the unknown elements of these treatments.

Another consideration related to informed consent is patients’ decision-making capacity, especially as PEIs are typically used for severe or treatment-resistant cases of depression (Beeker et al., 2017; Dunn et al., 2011; Mandarelli et al., 2018). A psychiatric diagnosis does not preclude intact decision-making capacity, but capacity may be compromised in some treatment-refractory cases. A survey on medical professionals’ attitudes about the unique ethical, legal, and policy questions that DBS presents revealed concern about the fidelity of the consent process (Ali et al., 2016). Similarly, Lawrence and colleagues interviewed psychiatric inpatients with TRD and found that they believed their depression interfered with their ability to effectively weigh risks and benefits of treatment options(Lawrence et al., 2019). Involuntary treatment usually occurs only in cases of emergency, much less often than popular culture depicts. Interestingly, Takamiya and colleagues found that 80% of patients surveyed reported overall satisfaction with and positive attitudes toward ECT regardless of whether voluntary consent was secured or if it was used on an involuntary basis (Takamiya et al., 2019).

Finally, neuroethicists—and bioethicists more generally—are increasingly focusing on social justice considerations, including how individuals with limited resources and lower socioeconomic status may be disproportionately affected, as poverty is correlated with higher burden of mental illness (Lake, 2017). Effectiveness is also entangled with principles of justice. For example, we may be wasting resources on treatments that are not bringing a benefit to patients.

Overall, our results highlight the importance of allocating resources towards the dissemination of information about structural aspects of care, including raising awareness of available mental health services and service providers, and information about the accessibility of services. Furthermore, this is the first study examining perceptions of a range of neurostimulation and neuromodulation interventions across several stakeholder groups. The few studies of barriers to the use of PEIs that do exist only consider ECT. Given that ECT has been controversial and is often misunderstood by patients and the public, generalizing these results to other PEIs may not be warranted and one cannot assume similar barriers are perceived across other types of PEIs. Prior to our study, there was no empirical evidence that there were similar perceived barriers. Our study therefore helps to form the bedrock of our understanding of PEIs by filling a gap in the literature.