Current ethical dilemmas experienced by therapists who use social media: A systematic review

Background: Despite the rise in therapist use of social media, current ethical guidelines have not provided clear guidance on how to use social media. Ethical guidelines commonly suggest therapists apply the same ethical principles advised for their offline work, without thorough consideration of differing potential ethical dilemmas online. Methods: A systematic review was conducted to explore the main ethical dilemmas encountered by therapists who use social media. Three databases were searched: PsycINFO, Web of Science and CINAHL Plus. A total of 52 studies were screened by their titles and abstracts, with a subsequent 36 studies analysed against inclusion and exclusion criteria. The remaining 14 studies were assessed using quality standards and included within the thematic synthesis. Findings: Data analysis conceptualised three themes: therapist searches risking the therapeutic relationship with subthemes of loss and trust and power imbalance; therapist defensiveness in response to client searches with subthemes of self-disclosure and self-protection; and societal normalisation of social media with subthemes of fa-miliarity and ease and blurred boundaries. Conclusion: The societal prevalence of social media use suggests it is now inevitable that therapists will experience some form

Psychotherapy (BACP; 2019) including specific practical suggestions for how to use social media and the implications of this work through to the British Health and Care Professionals Council (HCPC; that recommends adhering to the same ethical guidelines online as offline. The APA (2021) divides their guidelines into two categories: ethical and professional issues, such as considering the implications of searching for clients and colleagues to gain information; and education, training and professional development issues, such as educating supervisees on the appropriate use of social media.
The proposed update for HCPC guidelines scheduled for publication in September 2023 includes revised guidance on social media use, which relates specific standards to both personal and professional social media use. This includes ensuring content shared on social media is not discriminatory, which may hinder service users' access to services, and consideration that content shared is accurate and in line with government public health messages.
Providing guidance to therapists around social media use can be a challenging area to navigate. Detailed guidelines may be overly prescriptive and quickly outdated, whereas suggestions to use the same guidelines online as offline can be interpreted as being overly vague and difficult to translate into real-world contexts: for instance, are the steps a therapist might need to consider to manage risk and confidentiality on a social media platform the same as in person? Although there is some overlap in the ethical issues experienced in the therapy room and on social media, it is argued that social media brings additional manifestations of ethical dilemmas to consider (White & Hanley, 2022). As a consequence, therapists may work outside their competency online in terms of both the content that they share, as they may be perceived as a trustworthy representative of their field, and their understanding of how to use different social media platforms. Moreover, there are questions around responsibility for risk management when the general public who view the therapist's social media accounts are not their clients, in comparison with clear risk management policies for therapists working with clients.
Even prior to social media, studies have debated ethical dilemmas related to using the Internet more broadly, such as searching for clients on Google (Clinton et al., 2010). The increase in studies on ethical dilemmas from health practitioners' social media use clearly maps onto the start of the 2010s when social media use rapidly proliferated and new platforms were founded (Gabbard et al., 2011;Ginory et al., 2012). These studies collectively questioned how to maintain professional boundaries on social media, with psychiatrists viewing patient profiles on Facebook and receiving friend requests from patients. Jent et al. (2011) found doctors believed patient searches were not an invasion of privacy as the information was public online. Drude and Messer-Engel (2021) published recommendations for how to develop guidelines to regulate social media use by therapists, including the need for guidelines on areas such as confidentiality, informed consent, multiple relationships and working within their competency. The recommendations acknowledge the diversity and individuality of different ethical dilemmas experienced online; thus, a prescriptive approach cannot fully replace the use of professional judgement. However, whilst these recommendations have been made alongside a plethora of other studies providing guideline suggestions (Baier, 2019;Smith et al., 2021;Tutelman et al., 2018), education on how to manage ethical dilemmas on social media is not currently mandatory within therapy training programmes (White & Hanley, 2022). Furthermore, the primary author reflects on a lack of education on therapist use of social media whilst training as a counsellor and as a counselling psychologist in the UK, which inspired this research.
The vast majority of studies exploring ethical dilemmas on social media appear to be opinion pieces providing recommendations as above, or quantitative surveys. Khan and Loh (2021) provide one of the few qualitative studies interviewing Australian doctors to understand their intentions for using social media. The doctors reported some positive uses, such as the dissemination of health information and community support. However, they also reported barriers to social media being used more widely by healthcare professionals such as privacy and blurred professional boundaries. The doctors also reported concerns around poor-quality health information being shared online by other doctors, with the general public unlikely to fact-check information shared by doctors for accuracy and trustworthiness. There is a clear gap in the literature for further qualitative studies to explore ethical dilemmas experienced by other health professionals, particularly mental health professionals.

| Rationale and research question
Although there are multiple opinion pieces on therapist use of social media and recommendations for use, there appears to be a gap in the literature with no current systematic review on ethical dilemmas experienced by therapists on social media. As social media becomes

Implications for practice and policy
• This paper highlights the variety of ethical dilemmas experienced by therapists on social media, which could be incorporated into therapist training and ethical guidelines to prevent risk online to both therapists and the general public.
• The paper encourages self-reflection on how therapists feel when clients search for their therapist online.
• The paper questions therapist motivations for searching for their clients online and how this may impede the therapeutic relationship.
• As social media becomes increasingly integrated and normalised within society, ethical dilemmas around social media use will continue to grow. Thus, it is imperative that therapists have considered these issues and professional bodies have considered what might be construed as malpractice.  (Helsper & Eynon, 2010), the number of therapists using social media is likely to continue to grow and so the number of therapists experiencing ethical dilemmas online will similarly increase. Only through greater awareness of the variety of ethical dilemmas experienced by therapists on social media can further training and guidelines be provided to prevent risk online to both therapists and the general public.
The following research question will be considered in this systematic review: What does the existing literature tell us about the main ethical dilemmas encountered by therapists who use social media?

| ME THODOLOGY
A systematic review is a thorough review of literature conducted in a methodical manner, with the aim of synthesising the characteristics and findings of the included studies (Dempster, 2011;Higgins et al., 2019). Despite qualitative research previously being excluded from systematic reviews, it is now deemed appropriate to synthesise qualitative research and possible to combine both qualitative and quantitative research in one systematic review (Dixon-Woods et al., 2005;Dixon-Woods & Fitzpatrick, 2001).

| Search procedures
The aim of this search was to find studies that shared information on ethical dilemmas encountered by therapists who use social media.
Psychologists were selected as one type of therapist to focus on, with findings having implications for other mental health professionals. Three databases were searched: PsycINFO, Web of Science and CINAHL Plus. These databases were selected due to their psychological and sociological perspectives. The following key words were used within each database: Psychologist$ AND social media AND ethic$.
The reference lists of the remaining studies following the screening of titles and abstracts were also checked for further studies that may be relevant to this review. Many studies from the reference list search did not meet the inclusion criteria as they concerned other medical professionals, particularly doctors and psychiatrists.

| Eligibility criteria
The studies were assessed against the inclusion and exclusion criteria in Table 1.
Only research published after 2010 was included, as social media research quickly becomes outdated and less relevant to current ethical issues experienced on social media. However, this only led to the exclusion of one paper; thus, the research on therapist use of social media mirrors the rise of social media over the last two decades.
This review includes papers from any country due to the relatability of ethical dilemmas experienced by therapists on social media irrespective of geographical location. Originally, this study intended to only use qualitative research. However, there is a clear dearth of qualitative research in this area, only appearing in open-ended questions on mixed methods surveys. The decision was therefore made to include qualitative, quantitative and mixed methods research, which can offer greater understanding (Elliott et al., 1999).
Additionally, the inclusion criteria originally only included psychologists and trainee psychologists, as an example of a form of mental health professional for the systematic review due to requirements for the initial submission of this paper within a counselling psychology doctoral training programme assignment. However, the scope was widened to include studies that also interviewed other health practitioners, such as counsellors and psychotherapists, provided psychologists were still a large part of the sample and the focus was on the issues relevant to therapists.

| Data extraction
The studies were first screened by their title and abstract to determine their relevance to the research question. Duplicates from across the databases were removed at this stage. The remaining studies were reviewed according to the exclusion and inclusion criteria.

Inclusion criteria Exclusion criteria
Study sample: The sample must include applied psychologists, both qualified and those still in professional training, but can also include other mental health practitioners. Clients of psychologists. From any country rather than just the UK.
Study sample: Exclusively other practitioners, for example counsellors and psychiatrists.

| Quality criteria and assessment
The studies included had all been published in at least one online database. Quality was checked by the primary researcher using the Qualitative Assessment with Diverse Studies (QuADS) appraisal tool, designed specifically to quality-check multimethod studies in systematic reviews (Harrison et al., 2021). No studies considered the role of stakeholders in research design or conduct, which may reflect that this is more of a qualitative criterion whereas the studies included were predominantly quantitative or mixed methods surveys.
Despite this, every study scored highly on many criteria and hence were all included.

| Data synthesis
Thematic synthesis was used to analyse and synthesise data from the included studies (Thomas & Harden, 2008). The first stage was to inductively code the text based on meaning units within the discussion (Rennie, 1998 analyse the studies and store the codes and themes. The primary researcher (EW) followed the six recursive stages of thematic analysis outlined by Braun and Clarke (2013). There was a discussion of the coherence of the synthesis in supervision with the secondary author (Elliott et al., 1999).

| Ethical considerations
As this project focussed upon secondary information in the public domain, this review did not require a formal ethical review. It did, however, still comply with the University of Manchester (2022) Code of Good Research Conduct. The three databases were all screened on the same day by the primary researcher, to increase the likelihood of making similar screening decisions. The subjectivity of the research team must, however, be considered (Elliott et al., 1999).
The primary researcher is a trainee counselling psychologist who has received training on working with ethical dilemmas offline with clients, but not online. The second author is a Professor of Counselling Psychology with an interest in the interface between technology and therapy. Both are active social media users who follow a variety of health professionals. Neither researcher has personal experience of contact with clients online, and both have taken steps to increase their safety online, such as private personal accounts. Therefore, the themes created likely reflect some of the authors' own views and subjectivity, as is the nature of qualitative research (Braun & Clarke, 2021).

| Study flow
In total, 52 studies were found from the three databases. Nineteen studies were excluded following the initial checking of titles and abstracts, and a further six studies were removed due to duplication, leaving 27 studies. The titles and abstracts within the reference lists of these 27 studies were checked for further studies, with 10 studies added. The 37 studies were then compared against the inclusion and exclusion criteria, leaving a total of 14 studies. A PRISMA flow chart of this process is included in Figure 1. The characteristics of the 14 included studies are included in Table 2.

| Synthesis of data
Three main themes were generated from the thematic synthesis, as in Table 3. and so signifies a loss of trust. The words "invasive" and "violation" were used in multiple studies to refer to therapist searches for clients online, with subsequent detrimental impacts on the therapeutic relationship . This loss of trust seemed particularly pertinent with regard to the therapist not asking the client beforehand whether they could conduct the online search, as described by  below.
Thus, without first obtaining the client's permission through the process of informed consent, accessing a client's online information may be both inconsistent with these ethical ideals and a violation of the client's trust. (Lehavot et al., 2010, p. 165) The ethicality of client searches can be questioned further when therapists' searches are to "verify information" due to a distrust in what the client shares in session (Kolmes & Taube, 2014, p. 7). Additionally, even if the therapist intention is merely curiosity, they may stumble across information online that contradicts what is shared in session . Thus, this lack of trust works both ways where the therapist is unsure what to believe, and raising this incongruence with the client may result in the client losing trust in the therapist due to an invasive online search followed by a challenge in session: if informed consent was not obtained before the search and a discrepancy was found in client's insession reports and information obtained online, potential damage to the therapeutic alliance could result. (Harris & Robinson Kurpius, 2014, p. 17) Therefore, therapists searching for clients on social media can result in a loss of trust from both sides with a rupture in the therapeutic relationship.

Subtheme 2: Power imbalance
Lack of informed consent prior to therapist searches represents a clear power imbalance. Harris and Robinson Kurpius (2014) reported therapists not providing clients with the opportunity to consent or documenting the search, perhaps because they believed the client would find the search unethical and refuse consent.
Of the participants who conducted an online client search, more than 80% indicated that they never obtained client consent, did not document the search in the client's file, did not consider the possibility of having to breach confidentiality, and never discussed the benefits and drawbacks of the search with their client. (Harris & Robinson Kurpius, 2014, p. 17) DiLillo and Gale (2011) suggest therapists may reason with themselves that "What my client doesn't know won't hurt him/ her" to excuse searches for clients even when they know this is "ethically questionable" (p. 164). The therapist assumes the power in deciding F I G U R E 1 PRISMA flow diagram of the search process. Exploring areas of concern for psychologists using social media. Moreover, multiple papers raised the ethical dilemma of a client attempting to connect with the therapist, such as sending a friend request . Although the therapist may want to reject the request to maintain their privacy and professional boundaries, the decision to "accept, reject or ignore" is within their power, with each response "presenting unique challenges of their own in a psychotherapeutic relationship" (Van Allen & Roberts, 2011, p. 436).
Overall, therapists need to consider power dynamics when interacting with clients online, such as gaining client consent for searches and managing client expectations around not accepting friend requests.

| Therapist defensiveness in response to client searches
Subtheme 1: Self-disclosure Clients searching for their therapist online were frequently discussed within the papers, with the most common motive being curiosity to learn more about the therapist: Curiosity was the primary reason for seeking information, but a small group sought specific pieces of information that were withheld when the client asked the psychotherapist. (Kolmes & Taube, 2016, p. 152) This suggests an ethical issue in that if a therapist chooses not to self-disclose certain information with a client in session, the client may seek that information through social media. However, Anderson and Guyton (2013) acknowledged that therapists likely share more personal information on social media than they would be comfortable sharing with clients. "Full intentionality may be impossible" for self-disclosures, as the therapist cannot conduct a reflexive decisionmaking process of questioning what to disclose to a client if the client searches them online (Taylor et al., 2010, p. 157).
These findings also confirm clinician loss of control over disclosures of personal information online and the inability to confine these disclosures as one might do when making a deliberate disclosure that is carefully considered in terms of clinical impact on a specific client. (Kolmes & Taube, 2016, p. 151) There appears to be an incongruence between the therapist's "loss of control" in self-disclosure and the expectation that they are still responsible for the client's feelings in response to their findings. Moreover, their "reputation and credibility could potentially be harmed" (Van Allen & Roberts, 2011, p. 436) dependent on what the client finds, which may subsequently influence the therapeutic relationship and choice to remain in therapy. This subtheme indicates how therapist self-disclosure online may be viewed by clients without consent, and so therapists may need to discuss the impact this has on the client's feelings and perception of the therapist in session if the client raises this issue.

Subtheme 2: Self-protection
One of the studies reflected a pressure and responsibility which appears to be placed on the individual therapist to be responsible for managing their own behaviour and privacy online: individual practitioners must establish their own selfmonitoring strategies regarding online behavior. (Taylor et al., 2010, p. 158) Despite this individual responsibility, the therapists in the studies were unsure how to best maintain professional boundaries online due to a lack of training or discussion in supervision. The main behaviour change outlined across the papers was for therapists to "actively protect themselves by using privacy settings" The most extreme lengths proposed for therapists to protect their privacy are to refrain from social media use altogether.
However,  reflect that this is an unreasonable request for therapists given the current climate of social media use.
This subtheme exposes a discrepancy between many therapists perceiving clients searching their social media profiles as an invasion of privacy that they must protect themselves against, and yet in Theme 1, many therapists were willing to search for their clients online without deeming this as invasive or detrimental to the therapeutic relationship. Simply because so much information is now readily available and easily searchable online, this does not mean that obtaining information about our clients without their permission is ethically acceptable behavior. (Tunick et al., 2011, p. 445) Typically, supervision is the ideal opportunity for therapists to explore ethical issues. However, currently, this is a rare situation in that supervisees may have more online experience than supervisors and so not feel well supported or even capable of bringing this ethical issue . Education of supervisors and supervisees is essential as social media continues to evolve and be further integrated into daily life, with studies recognising therapist behaviour online will also likely change over time  DiLillo & Gale, 2011).

| Societal normalisation of social media
This subtheme emphasises how normalised social media has become in society; thus, therapists need to issue personal restraint to not search for clients and use supervision to reflect on new ethical issues presented by social media use.

Subtheme 2: Blurred boundaries
A recurrent subtheme was the blurring of boundaries on social media, which seemed more concerning and confusing than the maintenance of boundaries offline. Blurred boundaries alter the client's subsequent perception of the therapist's availability, as they have seemingly become easily accessible . Moreover,  questioned whether clients may then expect the therapist to check their profile between sessions and be responsive in caring for them in the same way as in the therapy room.
I think clinicians who read (client) blogs put the field at risk by setting a precedent that we ought to be responsible for that information. If we give people the idea that they can put things on their blogs and we will read it with the intent to intervene, what happens when they put a suicide note up at 2:30 a.m. and we do not see it? (Tunick et al., 2011, p. 445) TA B L E 3 Summary of themes and subthemes with example codes.

Themes Subthemes Example code
Therapist searches risking the therapeutic relationship 1. Loss of trust "Doing so violates a client's autonomy and has the potential to disrupt the element of trust that is essential in a therapeutic relationship." (Tunick et al., 2011, p. 445) 2. Power imbalance "The reported discrepancy between attitudes and actual search behaviors suggests that although trainees recognize that searches are ethically questionable, the ubiquity of these activities in their everyday lives may lead them to feel that client searches are of little consequence or are easily justified because of their anonymity (e.g., 'What my client doesn't know won't hurt him/her')." (DiLillo & Gale, 2011, p. 164) Therapist defensiveness in response to client searches 1. Self-disclosure "These findings also confirm clinician loss of control over disclosures of personal information online and the inability to confine these disclosures as one might do when making a deliberate disclosure that is carefully considered in terms of clinical impact on a specific client" (Kolmes & Taube, 2016, p. 151) 2. Self-protection "Most health care professionals who responded to the survey reported being aware of the ethical and professional implications involved and actively protect themselves by using privacy settings" (Anderson & Guyton, 2013, p. 125) Societal normalisation of social media 1. Familiarity and ease "just as doctoral students may think nothing of taking a few moments to learn something about a new social acquaintance online, so too may they quickly turn to the Internet as a source of information about their clients" (DiLillo & Gale, 2011, p. 164) 2. Blurred boundaries "some participants wrote that conducting a social media check of any kind runs the risk of compromising professional boundaries" (Vermeersch, 2022, p. 71) TA B L E 4 Development of codes and themes.

Descriptive theme Descriptive subthemes Codes
Therapist defensiveness in response to client searches (58) Changing online behaviour (7) Changing social media information (1) Come offline (1) Modifying online behaviour (1) Professional social media use (1) Representing the profession (1) Responsibility for self-monitoring (2) Privacy (19) Differing perspectives on privacy (8) Lack of privacy (1) Therapist responsibility to change social media privacy policies (1) Therapist responsibility to educate client on privacy (2) Use of privacy settings (17) Client finding therapist online (13) Client messaging/ friending therapist (5) Client messaging therapist (1) Friend request from client (3) Friending client (1) Client searching for therapist (8) Client searching for therapist online (3) Client searching for therapist's family (1) Client wanting information about the therapist (1) Lack of consent from the therapist (3) Self-disclosure (19) Consideration of what to share (3) Disclosure impacting therapist credibility or reputation (1) Disclosure to aid client decisions (1) Lack of control over disclosures (4) Not want clients to see information (1) Professional vs personal information (1) Therapist information shared online (1) Therapist self-disclosure (7) Societal normalisation of social media (46) Familiarity with social media (46) Familiarity and ease of using social media (27) Age of therapist (1) Ease of searching for clients (6) Evolution of social media (2) Familiarity with social media (4) Supervisor's lack of experience online (2) Normalisation of social media searching (8) Therapist restraint to not search (1) Unfamiliarity with technology (3) Boundaries online vs offline (19) Blurred boundaries (6) Inevitability of contact from clients (1) Pressure to friend supervisor (4) Therapist availability (4

Descriptive theme Descriptive subthemes Codes
Therapist searches risking the therapeutic relationship (121) Therapeutic relationship (79) Feelings after searching client (4) Discomfort at contact from clients online (2) Guilt, shame and embarrassment at searching (2) Therapist boundaries (8) Multiple relationships (4) Therapist availability (4) Power imbalance (25) Damaging therapeutic relationship (8) Power imbalance hiding knowledge (2) Lack of consent from client (15) Loss of therapist objectivity (6) Search not harming client (1) Therapist objectivity lost (3) Unanticipated impact on others (2) Trust (36) Confidentiality breached (1) Contradictory information online (5) Discussing searches with client (1) Discussion of clients on social media (1) Inaccurate information (2) Invasive searches (3) Losing trust (6) Not trusting client information (2) Therapist searching for clients online (42) Therapist searching for clients online (9) Purposefully searching clients (3) Searching for clients online (6) Ethics of searching (8) Adherence to ethical principles when searching (2) Documentation of searching clients (2) Still searching clients even if knowing not acceptable (2) Therapist disclosure of client information online (1) Unsure how to make ethical decisions (1) Reasons for searching (25) Client searches related to risk (10) Decision of why to search (5) Searching out of curiosity (6) Seeking information (1) Timings of searching (1) Verifying client information (1) Views on searching for clients (1) Lack of support (16) Lack of training (8) Not discussed searching in supervision (1) Seeking guidance (4) Supervisor's lack of experience (2) Use of supervision (1) Miscellaneous (14) Advertising on social media (2) Communication with other professionals (1) Contact with supervisor on social media (3) Impact on supervisory relationship (1) Lack of consent from supervisor (1) Online dating (1) Overlapping social circles with client (1) Payment on social media (1) Responding to risk (2) Supervisor searching supervisee (1) Unpaid contact with client (1) Note: The numbers in brackets represent the number of meaningful units within each code and theme. It is important to note that the frequency of meaningful units is not important within a qualitative approach; thus, this review focussed on the themes that best answered the research question rather than exclusively the themes with the highest frequency (Braun & Clarke, 2013). In total, there were 88 different meaning units. Contact from clients online, and thus the blurring of boundaries, was perceived as "inevitable (or at least not surprising)" (Asay & Lal, 2014, p. 108). This seems to instil a sense of defensiveness in therapists and a need to protect their information online as explored in the second theme, for seemingly when, not if, a boundary crossing occurs.
Psychologists should assume that their clients will search for their profiles.

| DISCUSS ION
The aim of this study was to understand what the current literature tells us about the main ethical dilemmas experienced by therapists who use social media. It provides a novel reflection on the ubiquitous nature of social media for therapists. Therefore, these findings are potentially relevant to all therapists, irrespective of whether they make use of social media within their practice as a mental health influencer or not. Through the systematic review process, three themes were created with two subthemes each, which will be discussed in the context of previous research. This section also considers the limitations of the review and suggestions for future research. The discussion ends with recommendations from the review for therapists who use social media that can be used within professional guidelines.
There was a degree of overlap between the themes, with a clear relationship between therapist searches risking the therapeutic relationship and therapist defensiveness in response to client searches.
Therapist searches of the client without consent negatively impacted the therapeutic relationship, but it can also be argued that client searches of the therapist may also impact the therapeutic relationship if the therapist then feels a pressure to protect themselves online from unwanted disclosures. This negative impact on the therapeutic relationship is substantiated in the previous literature regarding therapist self-disclosure online (Baier, 2019) and loss of trust from searching without consent (Cooper et al., 2019). The issue of multiple relationships online and in person with clients identified in this review is also corroborated by studies questioning whether this breaches the boundaries of the professional relationship (Demers & Sullivan, 2016;Lannin & Scott, 2013). Only through therapist openness and honesty can these ethical dilemmas be better understood and addressed, as the studies reflect a secrecy and shame in experiencing ethical dilemmas online.
Therapists appear to rarely discuss ethical issues on social media in supervision and so attempt to address these ethical dilemmas alone without support. This is reflected in the literature which encourages the use of supervision (Kolmes, 2012), whilst also recognising supervisors' lack of experience on social media may act as a deterrent for supervisees to share online ethical dilemmas, as identified in this review (Lannin & Scott, 2013). Supervisors also need to be aware of the role they play in potentially searching supervisees and clients without consent and the mistrust this may create (Kaslow et al., 2011).
It must be noted in this review that many therapists do not search for clients online. However, as social media use is increasingly integrated into daily life, and younger generations of clients and therapists grow up using social media, the likelihood of encountering some form of ethical dilemma increases. Younger therapists may find it more difficult to protect their privacy online, as they began using social media at a younger age so there is more information about them online (Kolmes, 2012;Nicholson, 2011). Due to social media algorithms, therapists and clients may also be unintentionally exposed to each other's accounts without intentional searching.
Moreover, this review identified that although therapists stated that they knew client searches were unethical, this did not deter them from occasional searches (DiLillo & Gale, 2011). Therefore, there is an importance in both gaining awareness of what the ethical dilemmas online may be and then having the knowledge to address these issues or minimise the risk of issues occurring, ideally incorporated within professional training (Campbell et al., 2016).

| Limitations of the review
A limitation of this review is that some of the studies included were from the early 2010s, when different social media platforms were more prevalent. Whilst many papers referred to therapists searching for clients on Facebook and vice versa, searches may now more likely be conducted on Instagram and TikTok, which are popular at the time of writing this review. Additionally,  combined the discussion of blogs and social networking together, whereas blogging appears much less common presently than other social media platforms. This is reflective of social media research in general, which quickly becomes outdated due to the rapid evolution of different platforms.
In contrast to the wealth of opinion pieces on ethical dilemmas experienced by therapists online, quantitative research appears to be much more focussed on a narrower range of ethical dilemmas.
Therefore, the review is not fully representative of other ethical dilemmas covered in opinion pieces that have not been researched quantitatively or qualitatively. For example, none of the studies in the review spoke of therapist use of social media for advertising, which is common in opinion pieces (Cederberg, 2017;Nicholson, 2011).

| Future research
The majority of the studies included in this review were quantitative or mixed methods online surveys.

| Recommendations
These themes summarise some of the ethical dilemmas experienced by therapists related to social media, which can be used to inform future guidelines. Whilst practitioners may not feel social media is relevant to their work, it is essential for ethical dilemmas around social media use to be included within continuing professional development activities or professional training for therapists. Such input will raise practitioners' awareness of potential issues and aid them in their decision-making when issues arise related to the increasingly ubiquitous world of social media. Consequently, reflexivity is needed from the therapist to consider how to manage their online presence in a way that will not be detrimental to the therapeutic relationship or their profession. Additionally, guidelines that are produced will need to be dynamic and regularly updated, as social media continues to evolve.
Multiple studies in the review referenced therapists wanting further specific guidance on how to use social media ethically; thus, there is a clear demand for guidelines. Such guidelines should explain the risks of searching for a client's social media accounts in the therapeutic relationship, with potential ruptures occurring due to the loss of trust. Furthermore, the importance of gaining consent needs to be included prior to searching for a client on social media, with the reasons for why the search is deemed necessary to the therapeutic work being reflected upon and the potential power imbalance considered. Future guidelines should also include how to increase therapist competency so that they feel capable of protecting their privacy online and to minimise potential unwanted self-disclosure that may also negatively impact the therapeutic relationship. The guidelines can recommend specific ways of discussing social media when working with clients, advocating that practitioners set clear boundaries from the start of their work around issues associated with different types of communication. Social media may increase the perceived accessibility of the therapist, so it may be key to discuss with clients how they can contact the therapist outside sessions. Importantly, therapists should bring these issues to supervision, with supervisors educating themselves on how best to support their supervisees.
Guidance on how to support supervisees with ethical dilemmas on social media could be included within supervision training.  (2013) Included as meets all inclusion criteria. Includes psychologists, psychiatrists and social workers.

PsycInfo
Asay and Lal (2014) Included as meets all inclusion criteria.

Web of Science
Baier (2019) Excluded as is an opinion piece.
From reference list Campbell et al. (2016) Excluded as is an opinion piece.

Web of Science
Cederberg (2017) Excluded as is an opinion piece.

Web of Science Cooper et al. (2019)
Excluded as is an opinion piece focussed on telepsychology more broadly.
CINAHL Plus  Included as meets all inclusion criteria. Includes psychologists and psychiatrists so decision taken to include.
From reference list Demers and Sullivan (2016) Excluded as is an opinion piece.

PsycInfo
Diamond and Whalen (2019) Excluded as is an opinion piece.
From reference list DiLillo and Gale (2011) Included as meets all inclusion criteria.
CINAHL Plus, Web of Science Drapeau et al. (2016) Excluded as is an opinion piece. Frankish et al. (2012) Excluded as only one psychologist in the study and mainly focussed on psychiatry.

Web of Science
From reference list  Included as meets all inclusion criteria. Includes psychologists and counsellors so decision taken to include.
From reference list Kaslow et al. (2011) Excluded as is an opinion piece.
PsycInfo Knapp et al. (2017) Excluded as is a book.
PsycInfo, Web of Science, CINAHL Plus Kolmes (2012) Excluded as is an opinion piece.
Web of Science  Included as meets all inclusion criteria. Includes psychologists and other mental health practitioners so decision taken to include.
Web of Science  Included as meets all inclusion criteria.
PsycInfo Lannin and Scott (2013) Excluded as is an opinion piece.
From reference list Lehavot (2009) Excluded as is an opinion piece and published prior to 2010.
From reference list  Included as meets all inclusion criteria.
From reference list Myers et al. (2012) Excluded as is an opinion piece.

CINAHL Plus
Nicholson (2011) Excluded as is an opinion piece.
Web of Science Nortje and Hoffmann (2015) Excluded as not specifically focussed on ethical issues related to social media.

PsycInfo
Palomares (2018) Excluded as chapter in a book.

PsycInfo
Pham (2014) Excluded as is an opinion piece.

Reference Included in thematic synthesis
Web of Science Smith et al. (2021) Excluded as is an opinion piece.
PsycInfo  Included as meets all inclusion criteria.
PsycInfo, Web of Science  Included as meets all inclusion criteria.
PsycInfo Tutelman et al. (2018) Excluded as is an opinion piece.

PsycInfo
Van Allen and Roberts (2011) Included as meets all inclusion criteria.
From reference list  Included as meets all inclusion criteria. Includes psychologists, social workers and counsellors so decision taken to include.
PsycInfo, Web of Science  Included as meets all inclusion criteria.
PsycInfo  Excluded as is an opinion piece and published prior to 2010.
From reference list Zur and Donner (2009) Excluded as is an opinion piece and published prior to 2010.