Opportunity Cost or Opportunity Lost: An Empirical Assessment of Ethical Concerns and Attitudes of EEG Neurofeedback Users

Abstract

Background:

Electroencephalography (EEG) neurofeedback is a type of biofeedback that purportedly teaches users how to control their brainwaves. Although neurofeedback is currently offered by thousands of providers worldwide, its provision is contested, as its effectiveness beyond a placebo effect is unproven. While scholars have voiced numerous ethical concerns about neurofeedback—regarding opportunity cost, physical and psychological harms, financial cost, and informed consent—to date these concerns have remained theoretical. This pilot study aimed to provide insights on whether these issues were supported by empirical data from the experiences of neurofeedback users.

Methods:

Semi-structured telephone interviews were conducted with individuals who had used EEG neurofeedback for themselves and/or for a child.

Results:

The majority of respondents (N = 36) were female (75%), white (92%), and of higher socioeconomic status relative to the U.S. population. Among adult users (n = 33), most (78.8%) resorted to neurofeedback after having tried other therapies and were satisfied with treatment (81.8%). The majority paid for neurofeedback out-of-pocket (72.7%) and considered it to be good value for money (84.8%). More than half (57.6%) considered neurofeedback to be a scientifically well-established therapy. However, of those, 78.9%were using neurofeedback for indications not adequately supported by scientific evidence.

Conclusion:

Concerns regarding opportunity cost, physical and psychological harms, and financial cost are not substantiated by our findings. Our results partially support concerns regarding insufficient understanding of limitations. This study underlines the disconnect between some of the theoretical concerns raised by scholars regarding the use of non-validated therapies and the lived experiences of users.

1. Introduction

Electroencephalographic (EEG) neurofeedback is a type of biofeedback that aims to train users to control the electrical activity of their brain [1]. It emerged as a promising tool to manipulate brainwave activity in the mid-1960s [2, 3] and has since been offered as a treatment for a wide variety of clinical indications, including attention-deficit/hyperactivity disorder (ADHD), post-traumatic stress disorder (PTSD), dementia, and depression [4, 5]. Although there are other types of neurofeedback, such as functional magnetic resonance imaging (fMRI) and functional near-infrared spectroscopy (fNIRS), which have shown some promise as treatments for certain clinical indications [6–8], these remain largely experimental and—as opposed to EEG neurofeedback—are not yet widely available to consumers. While currently there are no comprehensive reports regarding the precise size of the EEG neurofeedback market, the International Society for Neuroregulation and Research estimated that in 2017 there were more than 15,000 providers offering the technique worldwide [9]. Furthermore, it has been reported that there are approximately a dozen companies marketing at-home EEG neurofeedback devices [10].

Despite the widespread availability of EEG neurofeedback and the fact that there has been more research in this field compared to other types of neurofeedback, this modality remains controversial, as evidence regarding its effectiveness has been mixed [1, 5, 11–13]. To-date, there have been several studies in this field that present promising preliminary results for certain indications, such as ADHD and schizophrenia [14–17]. However, EEG neurofeedback research has been criticized for its lack of methodological rigor, as many studies utilize small sample sizes and few employ randomized controlled designs [5, 18]. Research that has applied such designs has been mostly limited to assessing the effects of neurofeedback for ADHD, and has indicated that neurofeedback may not be better than placebo in treating this condition [1, 5, 19, 20]. This includes a recent longitudinal multisite placebo-controlled study on EEG neurofeedback for ADHD, funded by the National Institute of Mental Health, which concluded that even though both the neurofeedback and control groups demonstrated significant improvements in attention, the results in the neurofeedback group were not superior to that of the control group [20]. Neurofeedback is not currently recommended as a treatment by the American Psychological Association (APA) or any other major professional medical society.

Given the lack of conclusive evidence to substantiate the effectiveness of neurofeedback, scientists and ethicists have raised a number of ethical concerns regarding the provision of the technique. First, individuals may opt for neurofeedback in place of more well-established therapeutic options, thereby resulting in a waste of time and resources [21, 22]. Second, even though neurofeedback is a low-risk, non-invasive therapy, there may be physical and psychological harms to users that may outweigh any benefits [22–24]. Third, neurofeedback may impose significant financial burden, as it has been reported that the cost of a treatment cycle of 30 to 40 sessions may range from $3,000 to $10,000 and is usually not covered by insurance [25, 26]. Fourth, it is unclear to what extent consumers have an accurate understanding of the limitations of the therapy, and whether they receive sufficient information before the beginning of their neurofeedback sessions that would allow them to make informed decisions about their healthcare management [11]. Fifth, neurofeedback services are frequently promoted for use on children [13, 27, 28], who are often considered to be more vulnerable than adults, because usually they cannot make their own healthcare decisions. As such, the risks brought using a therapy that is not well-validated are particularly heightened, as it is unclear whether these therapeutic options are in the best interest of children [29].

Notably, these ethical concerns parallel those raised by other non-validated therapies [30–32], and in particular those raised by many complementary and alternative medicine (CAM) techniques [33, 34]. Indeed, we have previously argued that EEG neurofeedback shares many common features with CAM [22] such as the absence of robust scientific data supporting its effectiveness, the lack of professional recommendations for its use by professional medical societies, and the marketing of services as a natural alternative to traditional healthcare. In the realm of CAM, scientists and ethicists have drawn attention to issues related to opportunity costs [35], potential harms [36], financial costs [34], inadequate informed consent processes [37] and use on children [38–40]. Yet, these critiques—both in the realm of CAM and EEG neurofeedback—are based on several implicit assumptions. First, it is often assumed that individuals may be foregoing more established therapies in order to pursue alternative ones. Second, it is hypothesized that they may be suffering physical and non-physical harms that are not outweighed by the benefits of the therapies. Third, individuals may be bearing significant financial cost. Fourth, they may not to be adequately informed about the potential and limitations of certain modalities. Finally, it is assumed that such treatments may not be in the best interest of children. However, the extent to which these assumptions are supported by evidence in the realm of EEG neurofeedback is unclear, as there has been no empirical work to assess whether the concerns expressed by scientists and ethicists are borne out by empirical data.

In this pilot study we aimed to fill this gap by conducting semi-structured telephone interviews with users of EEG neurofeedback. We sought to examine whether ethical concerns regarding opportunity cost, risk-benefit tradeoffs, financial costs, informed consent, and use on children were supported by empirical data. By obtaining preliminary insights into the experiences of users of EEG neurofeedback, we aimed to contribute to the literature examining user perspectives on CAM and other non-validated treatments.

2. Materials and Methods

2.1. Recruitment

The primary method of recruiting users of EEG neurofeedback was through social media. We searched both Facebook and Reddit for groups related to “neurofeedback” and included any group that had at least 100 members and had been active in the last year. Groups were excluded if they: (a) primarily used a language other than English; (b) were geared to neurofeedback providers and not users; or (c) were targeted to providers and users of a single company’s neurofeedback system. In total, seven Facebook groups and one Reddit forum met our inclusion criteria (see Additional File 1). Moderators of four Facebook groups and the one Reddit forum granted us permission to post recruitment messages. Between August and December 2020, we posted two messages to each group inviting current and past users of EEG neurofeedback to participate in a telephone interview (see Additional File 1 for recruitment message).

Due to a low yield from this recruitment method, we developed a secondary strategy that involved recruiting Twitter users who had posted public content indicating that they had used neurofeedback for themselves and/or for their children. We searched for the term “neurofeedback” across all public tweets posted between October and December 2020. Twitter users were included if: (a) the content of their tweet indicated that they had used neurofeedback for themselves and/or a child; and (b) their Twitter preferences allowed for direct messages. Between December 2020 and January 2021, we sent the 45 individuals who met our inclusion criteria a direct message inviting them to participate in our study.

All users interested in participating in our study were directed to an intake webpage where eligibility was assessed via a three-item survey that confirmed that individuals: had used neurofeedback services at least once (either for themselves or for their children); were over the age of 18; and were not providers of neurofeedback services. All participants who met our eligibility criteria were invited to schedule a telephone interview using an online booking system.

Individuals were offered a $20 Amazon e-gift card for completing the interview as compensation for their time and effort. After each interview, participants were invited to forward the information about our study to acquaintances who fulfilled our inclusion criteria and were offered an additional $10 Amazon e-gift card for each subject they helped us successfully recruit.

The protocol of this study was reviewed by the Institutional Review Board of the University of Pennsylvania in July 2020. All participants received an informed consent form via email prior to the interview, which was reviewed verbally before commencing each interview.

2.2. Interview guide

Interview questions were informed by a preliminary scoping review of literature related to EEG neurofeedback, CAM, and users of neurotechnologies. Questions assessed demographics, primary indications for use, and specific characteristics of usage (i.e., whether neurofeedback sessions had taken place in-clinic or by using an at-home device; and, where applicable, the reasons for discontinuing the treatment), as well as ethical issues related to neurofeedback (regarding opportunity and financial costs, informed consent, perceived benefits and unwanted effects), and attitudes towards the technique. The interview questions branched depending on whether participants reported that they personally had used neurofeedback, that their children had used neurofeedback, or both. The interview guide included both open- and closed-ended questions, which allowed the interviewer to cover a wide range of topics related to the experiences of participants while simultaneously offering them the chance to share their experiences in their own words (see Additional File 2 for the complete interview guide). Before commencing interviews, the interview guide was piloted within the research group to ensure reliability.

2.3. Analysis

Interviews were audio-recorded after obtaining verbal permission from participants. Open-ended responses were analyzed thematically in qualitative data analysis software Dedoose by two coders. The codebook for open-ended questions was developed by the two coders based on a thematic analysis of a subset of answers. Disagreements between coders were discussed until consensus was reached. Descriptive statistics were used to generate sample characteristics, and simple inferential statistics were used to explore significant differences between subgroups (SPSS, Version 26 [IBM]).

3. Results

3.1. Demographics, primary indications for use, and characteristics of usage

Of the 45 individuals who completed the eligibility questionnaire, 39 were eligible to participate in our study. Of those, 36 booked an appointment and completed a semi-structured telephone interview. All interviews were conducted by a single interviewer and took place between August 2020 and February 2021. In total, 26 participants were recruited through Facebook, seven through Twitter, and three via snowball sampling (see Additional File 1). During the data analysis stage, the inter-rater reliability between coders was 93.6%.

Detailed sociodemographic data are reported in Table 1. The vast majority of respondents (N=36) were white (91.7%), female (75.0%), and based in the United States (U.S.; 86.1%). The mean age of respondents was 44.9 years. Most were highly educated (approximately two-thirds had a bachelor’s degree or higher), and 80.5% reported earning over $50,000 annually.

Table 1.

Participants’ demographic characteristics

	N=36 (%)
Gender, n (%)
Male	8 (22.2)
Female	27 (75.0)
Non-binary	1 (2.8)
Age (mean years ± SD)	44.9 ± 11.5
Race/Ethnicity, n (%)
White	33 (91.7)
Other	3 (8.3)
Education, n (%)
High school or less	1 (2.8)
Some college	9 (25.0)
Associate’s degree	2 (5.6)
Bachelor’s degree	8 (22.2)
Master’s degree	11 (30.6)
Professional or doctoral degree	5 (13.9)
Household Income, n (%)
Less than $50,000	6 (16.7)
$50,000–100,000	13 (36.1)
$100,000–$150,000	8 (22.2)
$150,000+	8 (22.2)
Preferred not to answer	1 (2.8)
Employment Status, n (%)
Full-time	15 (41.7)
Part-time	2 (5.6)
Unemployed	4 (11.1)
Self-employed	8 (22.2)
Student	1 (2.8)
Homemaker	4 (11.1)
Retired	2 (5.6)
Country of Residence, n (%)
United States	31 (86.1)
Other	5 (13.9)

The largest group of respondents (n=33; 91.7%) used neurofeedback for themselves; this group is hereafter referred to as “adult users” of neurofeedback. Of those, one third (n=11) stated that both they and their children had used neurofeedback, and a handful of respondents (n=3; 8.3%) reported that only their children had used the modality. In total, there were 14 respondents who stated that their children had used neurofeedback (“child users”). Results for all adult users (n=33) are reported in the remainder of this section and in Sections 3.2.1–3.2.4; responses related to children are reported separately in Section 3.2.5.

The top three indications amongst adult users were anxiety, PTSD and ADHD (Table 2). Almost half of all participants used at-home devices. Amongst adult users of neurofeedback, some used at-home devices exclusively (i.e., without concurrently seeking in-clinic neurofeedback sessions), while others used at-home devices concurrently with in-clinic sessions, or only had in-clinic sessions (Table 2). The highest proportion of at-home device use was found among those who used neurofeedback both for themselves and for a child, with eight of 11 individuals reporting such usage.

Table 2.

Primary indications for neurofeedback use and characteristics of usage among adult users (n=33)

What is your primary purpose for using neurofeedback? (adult users only)*	n=33	%
Anxiety	17	51.5
ADHD	8	24.2
PTSD	8	24.2
Depression	4	12.1
Sleep disorders	4	12.1
Trauma	4	12.1
Head injury	3	9.1
To improve focus/concentration	3	9.1
Other**	4	12.1
Have your neurofeedback sessions taken place in a clinic, or have you used a device at home, or both? (adult users only)	n=33	%
Clinic	19	57.6
Device	5	15.2
Both	9	27.3
How did you first hear about neurofeedback? (adult users only)*	n=33	%
Word of mouth	14	42.4
Online	10	30.3
Recommendation from a healthcare provider†	9	27.3
Other‡	4	12.1

^*Respondents could select more than one option

^**Other indications included grief management, general wellness, meditation, and movement disorder

^†This category includes both CAM and mainstream healthcare providers

^‡Other includes research studies (n=2), news reports and advertisements

Almost half of adult users became aware of EEG neurofeedback through word of mouth, with others hearing about it online or from a healthcare provider (Table 2). Two-thirds of adult users (n=22; 66.7%) reported being current users of the technique; one-third (n=11; 33.3%) had stopped using neurofeedback at the time of the interview. When asked about the primary reason for stopping neurofeedback, participants cited high cost (n=7), unsatisfactory results (n=2), and the completion of neurofeedback treatment (n=2).

No statistically significant differences in any response were observed between subgroups stratified by the following: age; race; education; income; between users of at-home devices and those receiving neurofeedback during in-clinic sessions; between participants who use neurofeedback for themselves and those who use it for their child.

3.2. Ethical Issues

3.2.1. Opportunity cost

Among adult users (n=33), 78.8% (n=26) stated that they had tried other therapies before starting EEG neurofeedback, such as medication, cognitive behavior therapy, and other treatments, including CAM (Table 3). Some had been disappointed with the results of these therapies (n=7); as one participant stated: “I’ve probably tried greater than 15 different medications and they’ve had a lot of problems and side effects. I’ve also done expensive therapy and the majority of it has not been productive” (P24). At least four participants (12.1%) mentioned choosing neurofeedback after exhausting all of their options. One participant said: “I didn’t choose neurofeedback over the other ones, I chose it as the last one, as the last resort” (P32). Many participants (n=15) reported using neurofeedback as a complementary course of treatment: “I used neurofeedback in combination with other therapies. I’m also doing EMDR [Eye Movement Desensitization and Reprocessing], and I am doing talk therapy” (P35).

Table 3.

Opportunity Cost: Use of alternative therapeutic options among adult users (n=33)

Did you use alternative treatments or therapies for your primary purpose before starting neurofeedback?	n	%
Yes	26	78.8
No	7	21.2
If yes, what did you use?*	n=26	%
Medication	19	73.1
Cognitive behavior therapy	13	50.0
Other	13	50.0
If yes, did you feel that these alternatives were effective?*	n=26	%
Yes	3	11.5
Somewhat	11	42.3
No	4	15.4
Other	10	38.5
If yes, did you stop using these alternatives?*	n=26	%
I am still using it	15	57.7
I stopped due to lack of effectiveness	7	26.9
I stopped due to side effects	3	11.5
I stopped due to high cost	1	3.8
Other	4	15.4

^*Respondents could select more than one option

Among participants who did not try other therapies before starting neurofeedback (n=7; 21.2%), all were aware of the existence of better-established pharmacological alternatives for their condition. However, most of them (n=4) were unwilling to use medication. For example, one participant mentioned that she was pregnant and therefore could not follow a pharmacological treatment at the time. Another participant recalled: “I have family members that are drug addicts, so taking pills or anything like that […] I’m not going to do that, to try and fix something […] I don’t want to just rely on a pill that I have to take all the time to tell me feel a different way” (P9). Even among participants who were open to using medication, many mentioned choosing neurofeedback because it was a non-invasive, low-risk, and natural treatment.

3.2.2. Benefits and unwanted effects

To better understand how participants perceived the risk-benefit tradeoffs with regard to the results of EEG neurofeedback, we asked them to describe both its benefits as well as any unwanted effects. Regarding benefits, most adult users (n=25; 75.7%) reported experiencing symptom improvement, including improved focus and concentration, reduced stress and anxiety, better sleep, felt less depressed, and experienced less migraines. Others reported emotional benefits (n=20; 60.6%); for example, some stated that neurofeedback contributed to their emotional development and helped them feel like themselves again. In this regard, one participant described having a “stronger sense of who I am” (P1), and another claimed that “[neurofeedback] gave me myself back” (P27). Some reported experiencing performance improvement (n=7; 21.2%), such as memory improvement or increased concentration. Several reported interpersonal (n=4; 12.1%) benefits, such as improved relationships with their families and social circles, which they attributed to a general feeling of being more calm and less irritable. For example, one stated: “Before I was really standoffish and kind of kept to myself a lot. Now it’s way easier going up to people at work and talking to people and having discussions and I don’t feel anxious or nervous or anything like that around them” (P19).

The majority of adult users (n=26; 78.8%) also reported experiencing various unwanted effects. Physical side effects (reported by n=20; 60.6%) included headaches, nausea, sensory sensitivity, and vertigo. Others (n=10; 30.3%) reported emotional effects, for example, that neurofeedback brought up trauma history, anxiety, and/or irritability. One participant stated: “We’ve had one or two times where it just pushed me into a depression slump, like I don’t want to get out of bed, I don’t want to, you know, do anything” (P35).

Overall, adult users considered unwanted effects to be minor and temporary, as they usually abated after adjusting the neurofeedback protocol. For example, one participant recalled: “The providers I was working with were really good at checking in with me on those things, both in session and between sessions, so if I came back […] later for a session and I had had a negative thing, they would kind of play with tweaking frequency within the session” (P15). However, there were a few exceptions where participants considered unwanted effects to be more serious. For example, in one case, a participant reported experiencing severe anxiety for a prolonged period of time and described feeling “unable to function because of it” (P2).

Most participants reported having an overall positive experience with neurofeedback. The vast majority of adult neurofeedback users (n=27; 81.8%), when asked, stated that they were extremely or somewhat satisfied with the therapy, six participants (18.2%) stated that they were neither satisfied nor unsatisfied and no participant reported being extremely or somewhat dissatisfied. This indicates that for most the benefits of the therapy outweighed the unwanted effects.

3.2.3. Financial cost

Based on information reported by 23 adult users whose EEG neurofeedback sessions took place in a clinic, the cost of sessions ranged from $60 to $299, with an average price of $120.16. Participants described different payment models, with some providers offering a discount for bundled sessions (e.g., $2200 for 20 sessions) and reduced costs for regular users. Amongst adult users who reported using a device (n=14), a few (n=4) purchased a device at costs ranging from $4,700 to $6,700. The majority (n=9) rented a device, at costs ranging from $600 to $1,000 per month, with the average cost being $775. One participant was unsure of the details of the payment model.

Most participants (n=24; 72.7%) paid for neurofeedback treatment out-of-pocket, as it was not covered by their insurance plan. The high out-of-pocket cost of the modality was often cited as a concern by participants, when asked about the cost of the therapy and whether they considered it to be good value for money. For instance, one participant stated: “I was a little nervous about doing it because of the enormous costs we’d have to put up front. And for it to not work, that would be a little devastating” (P10). Despite its cost, the majority of respondents (n=28; 84.8%) strongly or somewhat agreed with the statement that “Neurofeedback provides good value for money.” To that effect, one participant stated: “No, it’s not [covered by insurance]. That’s unfortunate. But it was worth it” (P18).

3.2.4. Perceptions of Neurofeedback and Informed Consent Process

We probed participants’ perceptions of the effectiveness and limitations of EEG neurofeedback through a variety of questions. First, we inquired about a subjective sense of knowledge about EEG neurofeedback. The vast majority of participants (n=31; 93.9%) reported that they were well-informed about the risks and benefits of neurofeedback (Table 4). Next, we assessed participants’ views on certain issues that have been particularly controversial in the literature (namely, to what extent neurofeedback is a well-established treatment and whether its results could be based on placebo effect). Many participants (n=19; 57.6%) considered neurofeedback to be a scientifically well-established therapy and a similar number (n=21; 63.6%) did not believe that neurofeedback could be partly based on placebo effect. Interestingly, of those who believed that neurofeedback was scientifically well-established (n=19), the majority (78.9%; n=15) were using neurofeedback for conditions other than ADHD, which is the only indication for which neurofeedback has presented some evidence of effectiveness [1, 5, 41, 42].

Table 4.

Perceptions of neurofeedback: understanding of limitations of neurofeedback among adult users (n=33)

	n=33 (%)
Statements	Strongly disagree	Somewhat disagree	Neither agree nor disagree	Somewhat agree	Strongly agree
I consider myself well-informed about the risks and benefits of neurofeedback.	0 (0.0)	1 (3.0)	1 (3.0)	13 (39.4)	18 (54.5)
Neurofeedback is a scientifically well-established treatment.	1 (3.0)	1 (3.0)	12 (36.4)	11 (33.3)	8 (24.2)
The benefits of neurofeedback are partly based on placebo effect.	14 (42.4)	7 (21.2)	5 (15.2)	5 (15.2)	2 (6.1)

Furthermore, some participants expressed certain beliefs that are not commonly held by mainstream medicine regarding the mechanism of action and effectiveness of neurofeedback. For instance, a commonly cited reason (n=7; 21.2%) for using EEG neurofeedback was the perception of neurofeedback as targeting the brain directly by addressing the problem at “its root” and—as one participant put it, as opposed to medication, neurofeedback does— “not just mask the symptoms” (P25). In this regard, some believed that the benefits of neurofeedback would be long-lasting, since “it’s retraining your brain for long term use” (P9). Similarly, another participant reported: “It just seemed like going kind of straight to the source and bypassing the cognitive part of it, with the attempt to truly calm down more at the nervous system levels” (P20).

In addition, we asked a series of questions to examine the extent to which accurate information had been provided to participants during the informed consent process. Of the adult users who had in-clinic sessions (n=28), 20 reported having signed an informed consent form before the beginning of their treatment. Of those, 14 remembered having discussed risks and benefits of neurofeedback. Furthermore, only half (n=10) recalled having a conversation about therapeutic alternatives with their neurofeedback provider. Even though discussing the risks and benefits of a treatment as well as therapeutic alternatives is considered to be integral for individuals to make informed decisions about their healthcare management, some respondents indicated that they were already familiar with this information and considered such discussions unnecessary. For example, one participant mentioned that their neurofeedback provider discussing alternatives with them “would have been condescending because I’ve already tried so many things” (P36).

3.2.5. Use of neurofeedback in children

Of all those who reported that their children were using neurofeedback (n=14), half stated that their child was less than 12 years old. Some participants reported using neurofeedback for multiple purposes. The most common indication for use among children were ADHD (n=10); other primary indications included, among others, anxiety (n=3), concussion, migraine, and autism (n=1 each).

Responses regarding child users largely paralleled those of adult users with regard to potential opportunity cost, the benefits and unwanted effects of neurofeedback, as well as informed consent processes. As with adult users, the vast majority of those who used neurofeedback for a child (n=12) stated that they had tried other therapies before starting neurofeedback. One participant stated the following regarding her daughter: “She is on medication. She is in other therapies. This was an adjunct. And you know, it was low risk and I thought there was a chance it could help, so wanted to try it” (P29). Furthermore, there was a strong preference among some parents to avoid medication for their children. As one participant noted: “I have a lot of concerns about medication for [ADHD]. I wanted to basically exhaust every option before going down that road” (P20).

Most parents (n=11) were satisfied with the results of neurofeedback and reported different types of benefits their children experienced, including symptom improvement, performance improvement (such as better school performance), as well as emotional and interpersonal benefits. One participant mentioned that their child experienced “less frustration, less irritability. I’d say also maybe, [they became] more affectionate, kind of more open in that way” (P17). As with adult users, many participants reported that their children experienced some unwanted effects (n=8), including physical side effects (n=5), such as headaches and tics. Some also experienced emotional or interpersonal unwanted effects (n=4). In this regard, a participant stated: “That first week was a bit rough. I felt like [my children] were arguing more and were just a little bit more irritated” (P25). In most cases unwanted effects were deemed to be minor and short-lived.

In parallel with responses regarding adult users, parents of child users reported that information offered by neurofeedback providers during the informed consent process varied in terms of completeness. Of the ten participants whose children used neurofeedback and had contact with a neurofeedback provider, six reported having a discussion regarding both the risks and benefits of the therapy, while only two discussed mainstream alternative therapeutic options, such as medication.

3. Discussion

To our knowledge, this is the first study to examine the characteristics and experiences of users of EEG neurofeedback. While our sample is relatively small and our results may not be generalizable to the broader population of users, some clear sociodemographic patterns emerged. Namely, the typical user of EEG neurofeedback was a white female with higher educational attainment and income relative to the U.S. population [43] [44]. This same sociodemographic profile is common amongst users of CAM [45–48], such as those who seek acupuncture [49], and chiropractic treatments [50].

With respect to ethical concerns regarding opportunity cost, our findings indicate that the vast majority of participants had used other therapies before resorting to neurofeedback. Many participants used neurofeedback as a secondary option, after trying therapies that they considered to have been ineffective for their condition. Furthermore, neurofeedback was often used in conjunction with—and not instead of—better-established therapeutic alternatives. As such, it seems that the decision to use neurofeedback did not preclude participants from pursuing other better-established therapies. These findings are in line with studies demonstrating that individuals who use CAM do not necessarily reject mainstream medicine or use it as a substitute for other alternatives [51–53], but rather incorporate CAM into a therapeutic plan that mixes and matches different health care options [51, 54, 55]. However, the fact that some participants considered neurofeedback to be their last resort may be concerning, as it raises questions about whether they decided to try this modality while being in a state of despair and vulnerability. This may amplify concerns regarding whether the decision to try neurofeedback was autonomous and informed, or whether it could have been influenced by potentially exaggerated marketing claims [22].

Satisfaction with neurofeedback was found to be high; most participants reported experiencing various benefits and mostly minor and temporary unwanted effects. As such, ethical concerns about physical and non-physical harms were not shared by participants. At the same time, the wide range of benefits described by participants often went beyond relief of symptoms in a strict sense and extended to psychosocial improvements. Critiques of neurofeedback—and CAM writ large—tend to focus on issues of effectiveness when weighing the risks and benefits of these therapies [33, 34]. As a result, in the absence of robust data proving the effectiveness of a therapy, even a comparatively minor risk seems unacceptable [56]. However, users of CAM therapies often report improvements in their psychosocial wellbeing, which indicates that the benefits of the modality, as experienced by users, may be broader than symptom relief in a strict sense [54, 57]. This underlines the difference in perspective between scholars speculating about potential harms and the lived experience of users of these therapies. It is important to underscore, however, that both benefits and unwanted effects reported by our participants reflect their own perceptions of the results of neurofeedback treatment. In addition, given that most participants used neurofeedback in conjunction with other therapies, it is not possible to determine the extent to which reported effects can be attributed to neurofeedback.

The overall satisfaction of participants was also reflected in the view of the majority that neurofeedback provides good value for money, despite reporting that both neurofeedback sessions in a clinic and renting or purchasing a neurofeedback device came at a non-negligible, out-of-pocket cost. This highlights that perceptions of cost can be highly subjective and depend on the value that individuals assign to a certain treatment, as well as on their ability and willingness to cover it. As such, most respondents in our study considered neurofeedback to be valuable for their well-being and, therefore, found the cost acceptable. This indicates that ethical concerns regarding the financial burden imposed on neurofeedback (and CAM) users to some extent disregard the subjective element entailed in determining whether certain out-of-pocket costs may be acceptable to users.

The majority of participants considered EEG neurofeedback to be a scientifically well-established treatment, despite the lack of consensus in the scientific community. In addition, some participants expressed beliefs that are not commonly held by mainstream medicine (e.g., that neurofeedback targets the brain directly). Similarly, the majority of participants rejected the view that the benefits of neurofeedback may be based on placebo effect, even though most double-blind, placebo-controlled trials in this field indicate that the benefits of neurofeedback are not greater than those of sham [1, 5, 41, 42]. In this regard, our findings indicate that even when users are confident that they understand the risks and benefits of a therapy, there may still be nuances regarding the evidence-base and limitations, underlining the need for transparent communication during the informed consent process [11].

Participants reported variation in the types of information communicated by neurofeedback providers before the beginning of the treatment, ranging from extensive discussion (of risks, benefits, and therapeutic alternatives) to no discussion at all. However, studies in both clinical and research contexts have found that patient recall of information presented in the informed consent process tends to be inaccurate [58–62]. Thus, it is possible that more information was conveyed during the informed consent process but not recalled by participants.

Finally, results regarding child users were largely in line with those reported for adult users. In view of the high satisfaction rates among parents, the relatively minor unwanted effects, and the absence of opportunity cost for child users, our data do not raise major concerns regarding potential harms. However, the fact that half of the child users were reported to be under 12 years of age underlines their diminished autonomy. While parents are presumed to make decisions that are in the best interest of their children, as with other CAM therapies, parental autonomy may be constrained by overestimation of the effectiveness of neurofeedback and potential deficits in the informed consent process [39]. In view of the above, consulting a pediatrician or other healthcare provider could help parents ensure that neurofeedback is an appropriate course of treatment for their child, based on the child’s condition and specific circumstances.

Interestingly, no significant differences were reported between the answers of participants whose neurofeedback sessions took place in a clinic and those using at-home devices. This suggests that despite the differences in the contexts where neurofeedback was provided, the respective presence or absence of a provider, and the variations in costs, these factors do not seem to have played a significant role in participants’ satisfaction with the modality and their perception of unwanted effects and benefits of the therapy.

Our study has a number of limitations. First, as noted earlier, given that our sample was relatively small, our results may not be generalizable to neurofeedback users writ large. Second, our recruitment methods may have resulted in the over-representation of some users based on the groups from which they were recruited (e.g., one Facebook group was dedicated to parents of children using neurofeedback, and another was focused on neurodevelopmental trauma). In addition, our sample may have been biased towards neurofeedback users who tend to be active on social media. Third, as we did not interview child users directly, all data reported for children are based on reports from their parents. Finally, it is unclear whether those who opted-in our study differed from those who did not, potentially resulting in a nonresponse bias.

3. Conclusion

The results of this study underline the potential disconnect between some of the theoretical concerns raised by ethicists regarding the use of non-validated therapies, and the lived experiences of users. In principle, it is essential that ethicists interrogate the acceptability of such therapies, examine their ethical and social implications in order to raise awareness of these issues, and contribute to the mitigation of potential harms. However, many—although not all—of these speculative harms are value-laden and largely dependent on the subjective experiences of individuals. For this reason, it is important that the bioethics discussion is informed by empirical evidence. At the same time, it needs to be kept in mind that qualitative studies like the present one are not meant to provide definitive answers to these bioethical questions; they are merely one piece of the puzzle. Our aim is not to claim that certain concerns raised in the literature are negated by the experiences and beliefs of users, but rather to elucidate different perspectives and add another layer of consideration to the discussion. A better understanding of individuals’ perspectives and choices regarding their healthcare management can help us determine the real-world relevance of ethics concerns, and lead to pragmatic and evidence-based policy solutions.