Hot Topic Series; Issue #3
Ethical Considerations in the Proliferation of Large Online Therapy Platforms

Zac Kramer, MA; Dina Wirick, PhD; and the OPA Ethics Committee

This article is the third and final installation of our “hot topics” series, in which a controversial area in the mental health field is explored using the framework of the American Psychological Association’s (APA) Ethical Principles of Psychologists and Code of Conduct (hereby referred to as the “Ethics Code”) (APA, 2017). The hope of the OPA Ethics Committee is to inspire productive dialogue and promote useful inquiry about how to effectively approach these difficult topics, recognizing that our ability as a field to take collective ethical action requires curious, collaborative, and respectful dialogue.

Introduction
A 2022 cross-sectional study of around 70,000 individuals found that 12.8% of United States adults reported an unmet need for mental health counseling (Nagata et al., 2022). There were approximately 260 million adults in the U.S. at that time (Ogunwole et al., 2021), which means that, at the time of the study, roughly 33 million Americans were living without needed mental healthcare.

Given this finding, the proliferation of large online therapy platforms (such as BetterHelp and Talkspace) providing online mental healthcare would seem a welcome development, helping the allied mental health fields combat the overall national mental health crisis. As heartening as this development sounds, the relatively novel nature of the services that online therapy corporations provide––and the scale of their operations––introduce a host of potential ethical issues. While many large online therapy platforms (LOTPs) exist, this bulletin will focus specifically on BetterHelp, the largest of these LOTPs. BetterHelp (BH) has been subject to significant ethical scrutiny, primarily related to their confidentiality policies and other business practices. Despite the focus on BH, readers can consider how these issues may apply to the larger LOTP ecosystem and, perhaps even more broadly, to application- and web-based mental health services.

Application- and Web-Based Mental Health Services
A variety of mental health services are available over the internet, including self-paced cognitive-behavioral modules offered by organizations like Online-Therapy (www.online-therapy.com); mindfulness applications (“apps”) such as Headspace (www.headspace.com); artificial intelligence (AI)-generated mental health chatbots such as Earkick (www.earkick.com); and LOTPs such as BetterHelp. An in-depth review of the online mental health service landscape is beyond the scope of this article but, in general, these services have been evaluated quite positively in the literature. For example, a 2018 study found that internet-based cognitive-behavioral therapy (ICBT) and face-to-face therapy produced equivalent overall therapeutic outcomes (Carlbring et al., 2018). A systematic review of mindfulness apps found that they reduced depression in community samples, though results were mixed for improving anxiety and stress (O’Daffer et al., 2022). A 2022 study analyzing nearly 4,000 Google reviews of popular self-paced cognitive-behavioral and AI-powered apps found a high degree of consumer trust in these services, with a majority of user reviews attesting to reductions in negative affect, increases in positive affect, and/or greater self-insight (Shan et al., 2022). Interestingly, about a quarter of consumer reviews from this same study framed these self-paced and AI modalities as an alternative or complement to a live therapist (Shan et al., 2022). Finally, a recent Canadian study on user experience of 11 of the most popular mental health apps (including BetterHelp) found that these apps both empowered clients and increased trust in their providers (François et al., 2024).

Where is the Problem?
In the internet era, the allied mental health fields have made considerable advances in the effective, confidential, and ethical use of online/remote mental healthcare, including during the COVID-19 pandemic (Moudatsou et al., 2024; Wright & Caudill, 2020). Despite these research advances, remote therapy and mental health applications are still a relatively young technology, with applications like BetterHelp becoming more popular each year (Federal Trade Commission [FTC], 2023). Some experts believe that legislation and ethical guidance simply cannot keep up with the rapid development of internet technologies, creating the potential for unethical or poorly regulated use of these technologies (Wadwha, 2014; Wallach, 2015).

The Ethics of BetterHelp
The home page of BH displays the sheer magnitude of the services they provide. BH claims to contract with over 35,000 therapists and to have served over 4.6 million people (BetterHelp, 2024). Their site reports that BH has extended nearly 373 million “text messages, chats, phone, and video sessions” to its consumers, making it “the world’s largest therapy service” (BetterHelp, 2024). Moreover, according to one popular press article, an increasing number of college counseling centers are referring students to BH to meet demand for services, a trend that will need to be verified by a peer-reviewed study (Rosenbaum & Webb, 2023). As will be explored below, the scale of BH and its business practices introduce potential ethical concerns, specifically in the areas of confidentiality, marketing practices, and limited scientific evidence. Before psychologists endorse BH, recommend it to their clients, work for BH in a clinical capacity (all BH therapists are independent contractors; BetterHelp, 2024), or are employed by the company in managerial positions, they should consider whether their affiliation with BH aligns with our Ethics Code.

Confidentiality Violations
In its 2023 investigation of BH, the Federal Trade Commission (FTC) found over seven million instances of BH sharing consumers’ protected health information (PHI) with third-party advertising platforms like Facebook, Snapchat, and Pinterest (Fair, 2023). When users signed up for the BH platform, they were required to complete a questionnaire asking about sensitive health information such as mental health symptoms and psychiatric medications, and BH explicitly assured users that this information would only be shared with the client’s therapist (Fair, 2023). Instead, as the 2023 FTC complaint states, “From 2013 to 2020, [BH] continually broke these privacy promises, monetizing consumers’ health information to target them and others with advertisements for [BH]” (FTC, 2023, p. 2). The FTC report also found that BH had failed to train its employees on how to protect PHI and did not contractually limit how third parties could use patient information (Fair, 2023). The FTC banned BH from sharing consumer data and fined the company $7.8 million dollars (0.3% of parent company Teladoc Health’s reported 2023 revenue) in a settlement with consumers (Fair, 2023; “Teladoc,” 2024).

Insufficient privacy practices create potential ethical conflicts for psychologists who contract with BH, specifically related to Standard 3.11 (Psychological Services Delivered to or Through Organizations); Standard 4.01 (Maintaining Confidentiality); Standard 4.02c (Discussing the Limits of Confidentiality); and Standard 10.01 (Informed Consent to Therapy; APA, 2017). These ethical standards stipulate that psychologists have a “primary obligation” to protect clients’ health information and inform clients of potential privacy and confidentiality risks, including how larger organizations and third parties may use or share PHI (APA, 2017).

Questionable Marketing Practices
The scale of BH also introduces a multitude of ethical concerns around their marketing practices. Social media influencer marketing is a favored strategy among tech-savvy brands, with total spending on influencer marketing in 2024 expected to reach $5.6 billion in the United States and over $35 billion worldwide (“Influencer advertising,” 2024). The concept of influencer marketing is simple: brands invest in collaborations with social media personalities, who provide testimonials or advertising space for their sponsors’ products on streaming platforms like YouTube or Twitch, thereby delivering a targeted message to a brand’s desired audience (Deltombe & Bachmann, 2021). While this type of marketing is frequently used to sell beauty, nutrition, and sports products, the ethical implications are somewhat more complicated when it comes to selling mental health services (Deltombe & Bachmann, 2021).
According to a law blog on the Maastricht University website, BH has paid influencers to share “affiliate links,” i.e., links that provide influencers with commissions based on the number of individuals they refer to BH (Deltombe & Bachmann, 2021). In general, these brand partnerships are highly lucrative for influencers, who are likely incentivized to provide positive reviews for their sponsors in order to maintain their revenue streams (unless consumer backlash results in influencers distancing themselves from sponsors, as happened in the case of BH; Deltombe & Bachmann, 2021). Given these marketing practices, psychologists affiliated with BH may wish to consider whether contracting with BH is aligned with Ethics Code Principle A (Beneficence and Nonmaleficence), which encourages psychologists to be “alert to... financial,... organizational,... or political factors that might lead to misuse of their influence” (APA, 2017).

Where is the Evidence?
Generally speaking, research has found remote telehealth therapy sessions to have the same efficacy as in-person therapy (Lin et al., 2022). But several of BH’s features make it different from simply videoconferencing with a therapist who provides remote services. First, BH offers users the ability to text, live chat, phone call, and video call their therapist (BetterHelp, 2024). While each of these modalities by themselves has at least some support in the literature (e.g., see Hull & Mahan, 2017), BH combines these services into a multimodal mental healthcare medium that has limited precedent. There are only a few instances of scholarly support for the efficacy of multimodal psychotherapy platforms similar to BH (see e.g., Alvarez-Jimenez et al., 2020 and Mohr et al., 2010). In meta-analyses that evaluated the comparative efficacy of remote and in-person psychotherapy, studies that were included usually evaluated simple videoconferencing-based therapy and rarely included multimodal LOTPs (e.g., see Fernandez et al., 2021). BH also has other features that are not usually part of remote therapy. These include the ability to switch therapists an unlimited number of times at the press of a button as well as a fairly extensive questionnaire to match clients with therapists through BH’s proprietary algorithm (BetterHelp, 2024). The efficacy of these practices is not well established in the literature, and more research is needed to establish the potential benefits of multimodal features on platforms like BH.

Let’s assume for a moment that BH is an emerging treatment (i.e., a novel treatment without established efficacy). After all, BH has several features that differentiate it from ordinary videoconferencing-based therapy, and it operates at an unprecedented scale. Three Ethics Code standards would then apply (APA, 2017):

• 1. Standard 2.01c stipulates that “psychologists planning to provide services… involving… technologies new to them undertake relevant education, training, supervised experience, consultation, or study.”
• 2. Standard 2.01e states that “in emerging areas in which generally recognized standards for preparatory training do not yet exist, psychologists… take reasonable steps… to protect clients/patients… from harm.”
• 3. Finally, Standard 10.01b requires that, when psychologists obtain “informed consent for treatment for which generally recognized techniques and procedures have not been established, psychologists inform their clients/patients of the developing nature of the treatment” and “the potential risks involved.”

Imagine a psychologist contracting with BetterHelp who is asked by a client what the evidence base is for BH’s multimodal platform. What would that psychologist say? To date, there is a single study that specifically assesses the efficacy of BH as a mental health service. This study, by Marcelle et al. (2019), evaluated around 300 BH users experiencing symptoms of depression. It found that BH reduced depression symptoms (with moderate effect size) in participants who had used BH for at least 90 days (Marcelle et al., 2019). In the limitations section of the study, the authors acknowledged that the study lacked a randomly assigned control group and that their results may have been influenced by sample bias (Marcelle et al., 2019). What is more concerning is that the first and second authors of the study––graduate students at the time of publication––both reported conflicts of interest. The first author was a former consultant for BetterHelp; the second worked as a community and support manager for BetterHelp at the time of publication. This raises questions about any personal or professional incentives that may have impacted the results of this study, though no incentives were disclosed. These potential conflicts of interest call for additional research on the efficacy of BH to be conducted by an independent source.
How would BH answer the same question about the evidence base for their platform? One of the questions in the FAQ (frequently asked questions) section of BH’s website asks, “How can I be sure that this is an effective form of therapy?” (BetterHelp, 2024). In addition to citing the Marcelle et al. (2019) article above, BH states, “Our confidence in [our] platform comes primarily [emphasis added] from the feedback and testimonials we receive from users” (BetterHelp, 2024). Besides the Marcelle et al. (2019) study, a review of the scholarly literature did not turn up any studies specifically evaluating the efficacy of BH for mental health conditions. Despite this, BH’s website makes the claim that their platform can help users not only with depression but also with “anxiety, relationships, trauma, grief, and more” (BetterHelp, 2024).

BH is a technology company, not a psychologist, so they are not beholden to our Ethics Code. But psychologists who contract with or who are employed by BH are required, as stipulated by Standard 2.04 (Bases for Scientific and Professional Judgments), to base their work “upon established scientific and professional knowledge of the discipline” (APA, 2017). Psychologists contracting with BH who wish to provide evidence-based treatments (EBTs) through a multimodal platform which itself lacks robust scientific evidence will have to decide whether their clinical work is in keeping with the APA Ethics Code (2017). Note that not all BH therapists use multimodal features with their clients. Some BH therapists use only the videoconferencing feature with their clients, a practice more strongly supported in the scholarly literature (Fernandez et al., 2021). It is the combination of multiple modes of communication (i.e., video, phone, text, and live chat) that currently has limited scientific support. Should psychologists continue to provide services to clients through BH, Standard 10.01b (Informed Consent for Therapy) will also apply, obliging psychologists to inform their clients about the risks of BH, including that its multimodal features have not yet been shown to be effective for conditions other than depression (APA, 2017). Finally, Principle C: Integrity is relevant here as well, which encourages psychologists to “promote accuracy, honesty, and truthfulness in the science… and practice of psychology” (APA, 2017).

User Experience
With over 35,000 therapists, hundreds of millions of text messages exchanged, and several million clients served, there are bound to be instances of unethical conduct. In fact, hundreds of anecdotal user reports of grossly unethical and unprofessional behavior are available in user testimonials online. Note that these are anecdotal reports, almost all of them unverifiable. If true, many of these reports are concerning and serve as a call to researchers to study BH user experience. Some of these anecdotal reports (drawn from a popular YouTube video on the subject) of unethical/unprofessional behavior from BH therapists include: using the bathroom during session; operating out of non-private spaces; making popcorn during session; imposing religious values on clients; weight shaming; asking clients to detail their trauma narratives; using non-evidence-based wellness techniques; abruptly terminating clients without notifying them; and making homophobic remarks despite purporting to provide LGBTQ+ specialized therapy (Viola, 2024). But with over 35,000 therapists, none of whom BH provides supervision or consultation, BH risks creating an environment where unethical behavior may go unchecked and clients may find themselves in harmful therapeutic relationships without oversight (Lieberman, 2024). It is fairly standard practice for licensed independent contractors providing clinical services not to be provided with supervision or consultation. But given the scale of BH and the numerous anecdotal reports of therapist misconduct, psychologists should consider whether their affiliation with BH is in keeping with Ethics Code Principle B (Fidelity and Responsibility). Notably, the verbiage of Principle B states that, in addition to psychologists upholding their own professional standards, they are also “concerned about the ethical compliance of their colleagues’ … professional conduct” (APA, 2017).

Part of this potentially harmful environment includes therapist working conditions. While BH’s clinical and workflow guidelines for clinicians are not publicly available, anecdotal reports from licensed therapists who work or have worked for BH include claims that BH overworks and underpays its therapists (Lieberman, 2024). One therapist who claimed she worked for BH for over a year reported that BH encourages therapists to take on as many as 50 clients per week and that therapists who work 35+ client hours per week make twice as much per client hour as therapists who work only 5 to 10 client hours per week (Lieberman, 2024). If these anecdotal reports are accurate, these conditions may lead to lower-quality clinical work, which may not be in keeping with Ethics Code Principle A: Beneficence and Nonmaleficence (APA, 2017).

 
Conclusion
For psychologists who are aware of BetterHelp’s history of data-sharing practices, marketing strategies, and possible misrepresentation of the scholarly literature, continued affiliation with BH may put them in violation of the APA Ethics Code (2017). This includes BH psychologists who work directly with clients or are employed in non-clinical roles, as well as those psychologists who endorse BH or recommend it to their clients. One final standard is perhaps worth considering in light of the ethical concerns described throughout this article. That is Standard 1.03 (Conflicts Between Ethics and Organizational Demands), which states that, “if the demands of an organization with which psychologists are affiliated… are in conflict with this Ethics Code, psychologists clarify the nature of the conflict, make known their commitment to the Ethics Code, and take reasonable steps to resolve the conflict consistent with the… Ethics Code” (APA, 2017).

Large online therapy platforms (LOTPs) are not just remote therapy. The nature of their services and scale make them quite different from simple videoconferencing-based therapy. These are new technologies (BH was founded in 2013), and limited research exists examining potential ethical concerns with these services (BetterHelp, 2024). In fact, examining some of the ethical issues raised above required relying on some popular press articles and government reports, since scholarly research examining BH’s confidentiality, marketing, and other business practices is not available. BH’s sharing of PHI to third parties like Facebook, sponsorship deals with social media personalities, and sheer reach as part of a multi-billion dollar corporation raise unique concerns about the LOTP landscape. These ethical concerns have the potential to affect millions of consumers, just as the scale of LOTPs may potentially extend mental healthcare access to millions who need it. Finding the right balance will require the application of the APA Ethics Code––and ethics codes across the allied mental health fields––to new technologies, many of which the authors of our Ethics Code may not have anticipated or have not yet publicly addressed.

Readers should note that a new version of the APA Ethics Code is currently being developed by the APA Ethics Code Task Force (ECTF) (APA, 2024). One of the major planned revisions is around use of technology as it relates to the practice of psychology. For more information on these developments, see https://www.apa.org/ethics/task-force.