Virtual Reality exposure therapy in the treatment of public speaking anxiety and social anxiety disorder

Abstract

Virtual Reality (VR) has emerged as a promising tool in the treatment of anxiety disorders, particularly for exposure-based interventions. This literature review summarizes recent research on the efficacy of Virtual Reality Exposure Therapy (VRET) for addressing public speaking anxiety (PSA) and Social Anxiety Disorder (SAD). VRET demonstrates comparable efficacy to traditional in vivo exposure. While VRET exhibits relatively low drop-out rates, challenges related to immersion and sense of presence in the virtual environment remain to be addressed. Additionally, patients' preferences for VRET over in vivo exposure could enhance treatment adherence. The accessibility and cost-effectiveness of VRET render it a valuable therapeutic option, particularly for individuals with limited access to or low acceptance towards traditional therapy options. Self-guided VRET shows promise as an effective treatment approach, but its efficacy and safety necessitate further investigation. Ethical considerations pertaining to data protection and quality control are crucial when employing VRET applications. Careful implementation and monitoring are necessary to ensure patient privacy and safety. In conclusion, VRET holds significant promise as an effective and accessible treatment modality for SAD and PSA, offering new opportunities for prevention and management in the field of anxiety disorders. Continued research and advancements in VRET technology are expected to enhance its potential as a valuable tool in the mental health care landscape.

1. Introduction

Social Anxiety Disorder (SAD) is a frequent mental disorder and is associated with severely detrimental effects on the quality of life and the daily functioning of those affected (Kessler, 2003; Stein and Kean, 2000). It is characterized by an intense and persisting fear of social interaction or performance situations. The feared situations are avoided or endured under immense distress. Affected individuals often worry about potential embarrassment or humiliation, and frequently adopt dysfunctional safety behaviors (Fink et al., 2009; Schneier and Goldmark, 2015). SAD is estimated to have a lifetime prevalence of 5.5% in high-income countries and 3% in Western Europe (Stein et al., 2017), however, some studies have reported higher prevalence rates of up to 13% in Western countries (Brook and Schmidt, 2008). Additionally, the particular symptom characteristics of SAD often hinder affected individuals from seeking professional help, potentially resulting in an underestimation of the disorder's true prevalence (Dalrymple and Zimmerman, 2011; McGinn and Newman, 2013). Public speaking anxiety (PSA) constitutes a common subclinical subtype of SAD with prevalence estimates of around 30% (Stein et al., 1996). Individuals with PSA report excessive anxiety when speaking to larger audiences, leading to significant distress and impairment that does not extend to other social situations (Stein et al., 1996; Furmark et al., 2000; Lindner et al., 2019).

Psychotherapeutically, cognitive-behavioral therapy (CBT) represents the treatment of choice for SAD and PSA, with exposure exercises comprising the central element of CBT for SAD (Craske et al., 2008). However, despite strong empirical evidence supporting the efficacy of exposure therapy, it is often only sparingly realized in clinical practice (Neudeck et al., 2012). This may be due to several reasons, including restraints in time and available resources, which oftentimes result in exposure exercises being mostly assigned as homework without therapeutic supervision (McAleavey et al., 2014). Furthermore, therapists may be hesitant to incorporate exposure exercises in treatment (Becker et al., 2004; Farrell et al., 2013; Moritz et al., 2019), even after receiving specific training in this area (Trivasse et al., 2020). Most recently, the COVID-19 pandemic has furthermore caused a decrease in the application of exposure exercises in psychotherapy, mainly due to two reasons: First, pandemic-related restrictions such as social distancing or mask wearing mandates have resulted in an increased need for the remote delivery of therapeutic services, e.g., through video conferencing (Hong et al., 2021; World Health Organization, 2020). Second, the circumstances of a pandemic, such as hygienic measures or lockdown regulations, most often contradict the requirements of in vivo exposure to social situations. As a result, in-person exposure to social situations has become less feasible.

The existing treatment gap highlights the necessity for alternative, accessible and effective means of delivering exposure treatment for SAD. Amongst the increasing number of approaches to administer psychotherapeutic interventions with the help of novel technologies, few have been investigated as thoroughly as Virtual Reality (VR). VR enables individuals to confront fear-inducing stimuli within a digitally created environment, without the necessity of entering an aversive situation in real life. By employing computer-generated simulations, real-life movement tracking, and sophisticated modalities of displaying visual and auditory stimulation, VR allows patients to virtually experience the feared situation, and, consequently, perform exposure exercises.

The most common presentation modality for VR is the usage of VR goggles or head-mounted displays (HMDs), which usually consist of two separate LED displays, presenting a slightly different picture of a computer-generated simulation (or a 360° video) to each eye, and thus inducing 3D vision. Precise tracking of head movements makes it possible to adjust the presented perspective in VR, leading to a realistic and immersive impression of the perceived simulation. VR is more controllable and standardized than the traditional means of implementing exposure in sensu or in vivo. At the same time, it allows for a wide range of stimuli to be presented, such as performance situations or social interactions (Hartanto et al., 2014). Furthermore, VR exposure therapy (VRET) is more widely accepted among patients (Garcia-Palacios et al., 2001, 2007), and is more accessible, confidential, and resource-effective compared to traditional exposure in vivo (Carl et al., 2019; Emmelkamp, 2005; Miloff et al., 2016; Reeves et al., 2022). As a result, VR shows great potential as an alternate and efficient means of delivering exposure treatment for individuals with SAD and related phenotypes such as PSA.

In this review, we summarize the current state of research on the efficacy of VRET for SAD and PSA. The shared characteristics between the two are evident; nonetheless, recognizing the unique features of PSA treatment is essential for a thorough analysis of the efficacy of VRET.

Recent meta-analyses (Carl et al., 2019; Reeves et al., 2022; Morina et al., 2023; van Loenen et al., 2022; Horigome et al., 2020; Kampmann et al., 2016a) underscore the efficacy of conducting VRET for SAD with significant and enduring effect sizes (Hedges' g ranging from −0.86 to −1.14) over follow-up periods of up to 12 months. Additionally, VRET demonstrated substantial effects in addressing PSA, resulting in large and significant reductions (Hedges' g ranging from −1.39 to −1.46). Efficacies of VRET for both SAD and PSA appear comparable to in vivo exposure, emphasizing robust therapeutic potential of VRET. While the research cited above has laid a foundation for comprehending treatment efficacy, the present review aims to extend existing research by additionally providing a detailed summary of specific aspects such as treatment modalities, risks, and ethical considerations by consolidating findings from individual studies and discussing future clinical and research perspectives.

2. Material and methods

The literature search was performed independently by authors PDS and MAS using the PubMed and Web of Science databases by applying the following search terms:

(“Virtual Reality” OR “Virtual Audience”) AND (“Social anxiety” OR “Social Phobia” OR “Public speaking anxiety” OR “Performance Anxiety").

Additional studies were identified manually by searching reference lists of selected articles and pertinent reviews. Identified articles published until March 3rd, 2023 were screened based on title and abstract reading. Only peer-reviewed original research published in English, reporting treatment outcomes on samples with a primary diagnosis of social anxiety disorder according to DSM-IV/DSM-5 criteria or (subclinical) public speaking anxiety, and utilizing VR with a HMD for psychotherapeutic purposes was included.

3. Results

Table 1, Table 2 provide an overview of all articles on the treatment of PSA (Table 1) and SAD (Table 2) with VRET. The articles selected for our review encompass 20 studies, contributing enriching insights into the efficacy and modalities of VRET for social anxiety. Of the studies included, 45% (9 studies) compared VRET treatment with a waitlist control or no treatment condition, while 25% (5 studies) exclusively investigated VRET. Only four studies (20%) explicitly stated that VRET treatment was compared to in vivo exposure conditions. It has to be considered, however, that many studies employed active CBT comparison groups without clarifying to what extend exposure in vivo was part of the CBT treatment. Across studies, sample sizes varied from 10 to 97 participants. Utilized in 13 cases (65% of studies), the Liebowitz Social Anxiety Scale (LSAS; Liebowitz, 1987) constituted the most common instrument to assess symptoms of SAD and PSA as a primary or secondary outcome, directly followed by the Self-Statements During Public Speaking questionnaire (SSPS; Hofmann and DiBartolo, 2000), which was used in 6 cases (30% of the studies).

Table 1.

Overview of studies on VRET interventions for Public Speaking Anxiety.

Author	Country	N	Study Design	VR Intervention	Primary Outcome	Secondary Outcomes	Results
Denizci Nazligul et al. (2019)	Turkey	14	RCT	3 sessions VRET (8–10 min) vs. CBT	LSAS, IAS, SAAS, BFNE	SUD	Significant decreases in anxiety levels in both treatment groups.
Harris et al. (2002)	USA	17	RCT	4 sessions VRET (15 min) vs. Waitlist	LSAS, STAI, PRCS, ATPS, SUD	ATPS, HR, STAI	Significant decrease in PSA symptoms and heart rate after treatment.
Lindner et al. (2021)	Sweden	23	Single arm study	One session VRET (180 min)	PSAS	LSAS, BFNE, PHQ-9, GAD-7, BBQ, self-rated performance	Reduction in PSA, Increase in performance.
Lindner et al. (2019)	Sweden	50	RCT	One session therapist-led VRET (180 min) plus four weeks internet therapy vs. self-led VRET (180 min) plus four weeks internet therapy vs. Waitlist	PSAS	LSAS, BFNE, PHQ-9, BBQ	Significant decrease in PSA after treatment, maintained at 6- and 12-month follow-up assessments.
Lister et al. (2010)	Canada	15	RCT	4 sessions VRET vs. Waitlist	STAI, SPAI, SSPS	Skin conductance, HR	STAI-S and SSPS (neg) decrease after treatment. No change in SPAI, STAI-T or SSPS (pos) after treatment. HR and SC increase during VRET.
North et al. (1998)	USA	14	Controlled clinical trial	5 sessions VRET (10–20 min) vs. VR placebo	ATPS, SUD		Significant improvement after VRET.
Reeves et al. (2021)	United Kingdom	51	RCT	Four sessions VRET vs. VR-Control vs. no treatment	PSAS, LSAS	FNEB, IPQ	PSA reduction in both VR conditions, maintained at 10-week follow-up.
Safir et al. (2012)	Israel	49	RCT follow-up	Wallach et al. (2009) CBT with VRET vs. CBT	LSAS, SSPS, FNE, BAT		Effects maintained after 12 months.
Sarpourian et al. (2022)	Iran	30	Quasi-experimental study	One session VRET (30 min) vs. One session counselling (90 min)	PRPSA	LSAS, IPQ	SA not affected, reduced PSA in VRET condition.
Wallach et al. (2009)	Israel	88	RCT	CBT with 8 sessions VRET vs. CBT vs. Waitlist	LSAS, SSPS, FNE, BAT		Significant decrease of PSA after treatment. More dropout in CBT than VRET.
Wallach et al. (2011)	Israel	20 (plus 58 from Wallach et al. (2009))	RCT	12 session VRET vs. Cognitive therapy vs. CBT vs. Waitlist	LSAS, SSPS, FNE, BAT		Significant decrease of PSA after treatment. All active treatments of equivalent efficacy.

Legend toTable 1: PSA = Public Speaking Anxiety; NRT = Non-randomized Trial; RCT = Randomized Controlled Trial; VR = Virtual Reality; VRET = VR Exposure Therapy; CBT = Cognitive Behavioral Therapy; BAT = Behavioral Avoidance Test; PSAS = Public Speaking Anxiety Scale; LSAS = Liebowitz Social Anxiety Scale; PRPSA = Personal Report of Public Speaking Anxiety; IAS = Interaction Anxiety Scale; SAAS = Social Appearance Anxiety Scale; BFNE = Brief Fear of Negative Evaluation Scale; STAI = State Trait Anxiety Inventory; PRCS = Personal Report of Confidence as a Speaker; ATPS = Attitude Towards Public Speaking; SUD = Subjective Units of Distress; SPAI = Social Phobia and Anxiety Inventory; SSPS = Self Statements during Public Speaking; ATPS = Attitude Towards Public Speaking; FNE = Fear of Negative Evaluation; IPQ = Igroup Presence Questionnaire; SIAS = Social Interaction Anxiety Scale; PHQ-9 = Patient Health Questionnaire Depression; GAD-7 = Generalized Anxiety Disorder 7; HR = Heart Rate; BBQ = Brunnsviken Brief Quality of Life Scale.

Table 2.

List of all studies examining VRET interventions for Social Anxiety Disorder.

Author	Country	N	Study Design	VR Intervention	Primary Outcome	Secondary Outcomes	Results
Anderson et al. (2007)	USA	10	Single arm study	4 sessions prepared VRET	FNE, PRCS, PRCA-short, SSPS	BDI, STAI, CGI, CSQ-8, ImQ	Significant decreases in PSA after treatment and at follow-up.
Anderson et al. (2013)	USA	97	RCT	8 sessions VRET vs. exposure in vivo	FNE-Brief, PRCS, BAT	Speech duration, CGI, WAI, Homework compliance, CSQ-8, DSM-Diagnosis of SAD at FUP	Significant improvements of SAD measures in both treatment groups.
Anderson et al. (2017)	USA	28	RCT follow-up	8 sessions VRET vs. exposure in vivo	PRCS, BFNE	SCID, PGI, BAT	54.2% remission rate, 68% “very much improved”/“much improved”.
Anderson et al. (2005)	USA	10	Single arm study	4 sessions VRET	PRCS, SSPS, PRCA		Significant decreases in all self-report measures after treatment and at follow-up, but unchanged willingness to public speaking.
Bouchard et al. (2017)	Canada	59	RCT	CBT with 8 sessions VRET (20–30 min) vs. CBT with exposure in vivo vs. waitlist	LSAS	SPS, SIAS, FNE, BAT, BDI-II, SPRS, SSQ, PQ, GPQ	Significant improvements for both treatments in SAD symptoms and secondary outcomes, maintained at follow-up. VRET more effective than exposure in vivo.
Gebara et al. (2016)	Brasil	21	Single arm study	up to 12 sessions VRET (50 min)	LSAS, CGI, SF-26, SUD, Sheehan Disability Scale, SAS, ATQ30, DAS	Scale for Measuring Therapy Sessions (individually created for this study), BDI	Significant improvements in SAD symptoms and all other outcomes after treatment and at follow-up.
Geraets et al. (2019)	Netherlands	15	Single arm study	CBT with 14 sessions VRET (40 min)	SIAS, Paranoid Thought Scales	BDI II, Manchester Short Assessment of Quality of Life	Significant symptom reduction after treatment and at follow-up.
Kampmann et al. (2016b)	Netherlands	60	RCT	7 sessions VRET (60 min) vs. exposure in vivo vs. waitlist	LSAS, FNE-B	DASS, speech duration, PDBQ, EUROHIS-QOL	Significant improvements for both treatments in SAD symptoms and other outcomes; Exposure in vivo was superior to VRET regarding size, broadness, and stability of effects.
Klinger et al. (2005)	France	36	Controlled clinical trial	12 sessions VRET (20 min) vs. CBT (120 min)	LSAS	SCIA, HADS, SDS, CGI	Significant reduction in LSAS scores after treatment, equivalent in both treatment conditions.
Robillard et al. (2010)	Canada	45	RCT	16 sessions CBT with VRET vs. CBT with exposure in vivo vs. waitlist	FNE, LSAS, SPS, ASC, BDI, STAI trait, Self-efficacy single item		Significant reduction of anxiety after treatment.
Stefaniak et al. (2022)	Poland	73	Clinical trial	12 sessions CBT with VRET vs. self-administered VRET vs. CBT	LSAS	CGI, PGI, BDI, SSQ, SUD	Significant LSAS reduction after treatment. CBT and CBT + VRET exceeded self-administered VRET.

Abbreviations: SAD = Social Anxiety Disorder; PD = Panic Disorder; PSA = Public Speaking Anxiety; LSAS = Liebowitz Social Anxiety Scale; RCT = Randomized Controlled Trial; VRET = Virtual Reality Exposure Therapy; CBT = Cognitive Behavioral Therapy; FNE = Fear of Negative Evaluation; PRCS = Personal Report of Confidence as a Speaker; PRCA = Personal Report of Communication Apprehension; SSPS = Self Statements during Public Speaking; BAT = Behavioral Avoidance Test; BFNE = Brief Fear of Negative Evaluation Scale; CGI = Clinical Global Impression; SF-26 = Short Form Health Survey 26; SUD = Subjective Units of Distress; SAS = Social Anxiety Scale; ATQ30 = Automatic Thought Questionnaire 30; DAS = Dysfunctional Attitude Scale; SIAS = Social Interaction Anxiety Scale; FNE-B = Fear of Negative Evaluation Scale Brief; BAI = Beck Anxiety Inventory; STAI = State Trait Anxiety Inventory; ISS = Internalized Shame Scale; PERS = Post-Event Rumination Scale; SPS = Social Phobia Scale; HADS = Hospital Anxiety and Depression Scale; ASC = Appraisal of Social Concerns; BDI = Beck Depression Inventory; CSQ-8 = Client Satisfaction Questionnaire; lmQ = Immersion Questionnaire; WAI = Working Alliance Inventory; FUP = Follow-up Point; SCID = Structured Clinical Interview for DSM-disorders; PGI = Patient Global Impression Scale; SPRS = Social Performance Rating Scale; SSQ = Simulator Sickness Questionnaire; PQ = Presence Questionnaire; GPQ = Gatineau Presence Questionnaire; DASS = Depression-Angst-Stress-Skalen; PDBQ = Personality Disorder Belief Questionnaire; EUROHIS-QOL = Eurohis Quality of Life Scale; SCIA = Questionnaire of Social Context Inducing Anxiety; SDS = Social Desirability Scale; ABS-II = Attitudes and Beliefs Scale II; ITQ = Immersive Tendencies Questionnaire; RQ = Rumination Questionnaire.

3.1. VRET for public speaking anxiety

Several studies have probed the effectiveness of single-session VRET for subclinical PSA. An RCT (Lindner et al., 2019) examined the effects of therapist-led and self-led one-session VRET followed by a four-week, structured online-administered VR to in vivo transition program in a total of n = 50 participants with PSA compared to a waitlist control condition. In a three-hour session, participants received psychoeducative information and performed several speech tasks of increasing difficulty in front of a recorded audience in VR. Both therapist- and self-led treatments resulted in significant reductions in self-reported PSA symptoms in comparison to the waitlist condition. Comparing both active interventions directly revealed better outcomes at both six- and 12-month follow-up if the intervention was led by a therapist as opposed to the self-guided intervention. Notably, performance of in vivo exposure exercises during the post-VR transition period was associated with higher symptom decreases (Lindner et al., 2019). Applying an adapted version of the same protocol in routine clinical care, a decrease in self-rated PSA was observed in a single-arm design with n = 23 participants with PSA. Effects remained stable over a three-month follow-up period (Lindner et al., 2021). In a sample of n = 30 students with self-reported speaking anxiety, Sarpourian et al. (2022) compared one 30-min session of VRET implementing speech tasks in a virtual classroom (n = 15) to a 90-min counselling session (n = 15). Self-rated PSA significantly decreased in both conditions, while, however, reductions in self-rated social anxiety did not reach statistical significance.

In an RCT comparing VRET to CBT with exposure in sensu (Wallach et al., 2009), both active treatments were conducted over a span of 12 sessions. Participants in both groups initially received an identical treatment protocol comprising psychoeducative information and cognitive restructuring, while the behavioral component of treatment differed, with n = 28 participants with PSA receiving exposure in VR, and n = 30 participants with PSA receiving imaginal exposure. Both interventions led to a significant reduction in PSA symptoms, and they demonstrated comparable efficacy. Furthermore, both VRET and CBT with imaginal exposure were found to be superior to the outcomes observed in a waitlist control condition (n = 30) who did not receive any active treatment (Wallach et al., 2009). A follow-up analysis demonstrated that the effects were maintained after 12 months (Safir et al., 2012). In a subsequent follow-up study building upon the CBT with exposure in sensu and waitlist control data reported by Wallach et al. (2009), n = 20 newly recruited participants with PSA received either VRET (n = 10) or cognitive therapy without exposure (n = 10). Both treatment conditions were effective in reducing PSA symptoms, and no significant differences were found between groups. Furthermore, when comparing the outcomes to the data previously reported by Wallach et al. (2009), both VRET and cognitive therapy without exposure showed similar effectiveness to CBT with in sensu exposure. Additionally, both VRET and cognitive therapy without exposure were found to be superior to an untreated waitlist control condition (Wallach et al., 2011). In an RCT applying 360° video recordings, n = 17 participants with high PSA received VRET with gradual exposure in front of a virtual audience, while n = 16 participants were exposed to an empty virtual room of, however, varying size, and n = 18 participants received no treatment. Participants in the treatment and active control groups received four weekly VR sessions. A reduction in self-reported PSA symptoms was observed in both the VRET and the VR control condition without a virtual audience present. At a 10-week follow-up, effects were maintained compared to the group receiving no treatment, while no differences were observed between the VR groups (Reeves et al., 2021). In an RCT using a 3D shutter glass system, n = 9 participants with PSA received four sessions of VRET consisting of a reading task in front of a virtual audience. Compared to a waitlist control group (n = 6), post-treatment decreases in state anxiety and negative cognitions concerning public speaking decreased significantly. However, no differences were observed regarding symptoms of social anxiety, trait anxiety, and positive cognitions on public speaking (Lister et al., 2010). In a small-scale study, Denizci Nazligul et al. (2019) examined the effects of three sessions of therapist-led VRET in a small sample of n = 7 students with PSA compared to a PSA control group (n = 7) receiving CBT-based group psychoeducation. VRET consisted of speech tasks in front of a virtual audience of increasing size. Both groups showed significant reductions in anxiety measures at post-treatment, which did not differ between treatment conditions. In a small-scale study, n = 8 students with PSA received four weekly sessions à 15 min of public speaking VRET. In comparison to n = 6 participants in a waitlist control group not receiving any immediate intervention, VRET was found to be superior to waiting in reducing self-reported PSA (Harris et al., 2002). North et al. (1998) compared five sessions à ten to 20 min of public speaking VRET in n = 6 psychology students to n = 8 students who did not receive active treatment. The results revealed significant improvements in participants' attitudes towards public speaking as well as subjective distress during public speaking situations after the VRET intervention.

3.2. VRET for social anxiety disorder

In a small clinical study of n = 10 patients with either SAD (n = 8) or panic disorder with agoraphobia (n = 2) with a predominant fear of public speaking (Anderson et al., 2005), four sessions of anxiety management training followed by four sessions of VRET entailing speech tasks in front of a virtual audience were administered. Significant decreases in self-reported public-speaking fears were observed and found to be stable at a three-month follow-up assessment (Anderson et al., 2005). In a subsequent single-arm study by Anderson et al. (2007), n = 10 SAD patients received four sessions of VRET following a four-session computerized, psychologist-led self-help program. Public speaking fears decreased significantly post-treatment, and the effects were stably maintained at a three-month follow-up. In a subsequent RCT in n = 97 SAD patients with predominant fear of public speaking, Anderson et al. (2013) compared the VRET protocol applied in Anderson et al. (2005) with group exposure therapy and a waitlist control group. Both exposure-based treatments led to significant symptom reductions compared to the waitlist control condition, while both active treatments did not differ in regards of treatment efficacy. Furthermore, in a follow-up study involving n = 28 patients of the same trial (Anderson et al., 2017), treatment effects were maintained after four to six years, with more than half of the patients (54.2%) achieving complete remission. Another RCT in patients with SAD compared 14-session CBT with either eight sessions of VRET (n = 17) or exposure in vivo (n = 22) to a waitlist control condition not receiving active treatment (n = 20; Bouchard et al., 2017). Again, both exposure-based therapies were observed to be effective and significantly reduced self-reported SAD symptoms compared to the waitlist condition. While both exposure conditions were of comparable efficacy regarding most outcome measures, VRET exceeded the in vivo condition in the primary outcome measure, with significantly greater symptom reductions in patients receiving exposure in virtuo. Furthermore, the effects were maintained at a six-month follow-up time point (Bouchard et al., 2017). Another RCT compared CBT with VRET to CBT alone, as well as to self-administered VRET in a total of n = 73 SAD patients (Stefaniak et al., 2022). In both CBT conditions, patients received six sessions comprising psychoeducation and cognitive therapy, followed by either six sessions of virtual exposure (n = 29) or imaginal exposure (n = 25). Self-administered VRET (n = 19) was performed in a home setting without therapist contact. SAD symptoms significantly decreased in both the CBT + VRET and CBT conditions, and these improvements were greater than in the self-administered VRET group. Klinger et al. (2005) assigned n = 36 SAD patients to twelve weekly sessions of either VRET (n = 18) or group CBT with exposure (n = 18). Both conditions significantly reduced self-reported SAD symptoms and yielded similar levels of efficacy following treatment. In a sample of n = 45 SAD patients, Robillard et al. (2010) compared 16 sessions of CBT with VRET (n = 14) to CBT with exposure in vivo (N = 16) and to a waitlist control condition (n = 15). Significant symptom improvements were found in both patients receiving CBT with VRET and those receiving CBT with exposure in vivo. Both active treatment conditions did not differ from each other, but were superior to the waitlist control condition. In a single-arm study, Gebara et al. (2016) examined the effects of up to twelve sessions of VRET (with an average of seven sessions completed before scenarios no longer elicited fear) following one session of psychoeducation in n = 21 patients with SAD. Significant symptom improvements following VRET were observed, and effects were maintained at a six-month follow-up. In a small feasibility study, n = 15 patients with generalized SAD received two sessions of case conceptualization followed by up to 14 40-min sessions of VRET. SAD symptoms, as well as secondary symptoms of depression, decreased significantly after treatment, and effects were maintained at a six-month follow-up (Geraets et al., 2019).

In contrast to the studies reviewed above, a study compared the effectiveness of ten sessions of VRET entailing seven virtual exposure sessions in 20 patients with SAD to ten-session CBT with seven exposure sessions in vivo (n = 20) and to a waitlist control group (n = 20). Both the VRET condition and the in vivo exposure condition led to significant symptom improvements compared to the waitlist control group. However, notably, the in vivo exposure was found to be superior to the VRET condition in terms of reducing symptoms of social anxiety (Kampmann et al., 2016b).

4. Discussion

After systematically exploring the pertinent literature and synthesizing the resulting available studies on VRET for SAD and public speaking anxiety, this section delves into the interpretation and implications of the findings. Here, we assess treatment efficacy, accessibility, mode of delivery, variations in VR scenarios and discuss the role of therapeutic alliance as well as the promises and pitfalls within the evolving realm of VR technology, including ethical considerations, thus offering a perspective on the current landscape and future possibilities of VRET for SAD and PSA.

4.1. Treatment efficacy

Several meta-analyses have addressed the efficacy of VRET. Effect sizes for VRET compared to waitlist control conditions are estimated to be large (Hedge's g between 0.86 and 1.46) for SAD (Reeves et al., 2022; Morina et al., 2023; Horigome et al., 2020; Kampmann et al., 2016a) and PSA (Reeves et al., 2022), and moderate to large (Hedge's g between 0.49 and 0.78) for anxiety disorders in general (Carl et al., 2019; van Loenen et al., 2022). Compared to active control conditions, meta-analyses have revealed no significant difference (p values between 0.5 and 0.3; Morina et al., 2023; van Loenen et al., 2022; Kampmann et al., 2016a), indicating comparable efficacy of VRET to CBT with in vivo exposure. Moreover, the positive effects of VRET appear to be maintained over extended periods, with follow-up studies showing sustained benefits for up to six years (Anderson et al., 2017).

Treatment outcome studies commonly rely on self-rated questionnaires to measure treatment effects, which, consequently, is reflected in the effect sizes reported in meta-analyses. In the studies reviewed above, the LSAS (Liebowitz, 1987) emerges as the most frequently utilized outcome measure to assess symptom severity in both SAD and PSA, closely followed by the SSPS (Hofmann and DiBartolo, 2000). Despite this consistency, a broad variety of self-report questionnaires have been employed to measure clinical symptoms across studies (see Table 1, Table 2 for details). Observer-rated assessments, however, are used less frequently. However, their inclusion is recommended, as self-ratings are typically subject to biases and misinterpretations (Glass et al., 1987; Möller, 2009). Clinicians' ratings also show stronger relationships with objective measures such as heart rate variability (Ham et al., 2023), suggesting a potentially more reliable assessment. Furthermore, in clinical practice, the treating clinician's impression plays a significant role regarding treatment selection and duration, as well as the assessment of diagnosis or symptom severity. Implementing clinicians' ratings more consistently into clinical research would therefor increase external validity of the results and facilitate their translation to clinical practice.

While self- and clinician-rated instruments constitute the standard outcome assessment tools in both treatment research and clinical practice, they are limited by their subjectivity and may be susceptible to several biases. Including additional outcome measures such as physiological recordings alongside psychological assessments may increase objectivity and comparability across samples and experimenters. Although in some studies, physiological outcome measures are recorded, in the analysis they are often neglected or relegated to serve the purpose of a “manipulation check” (e.g., measuring heart rate to ensure that fear was experienced during exposure; cf. Lister et al., 2010). An increased adoption of behavioral measures to gauge the avoidance component of fear, which plays a central role in anxiety disorders and their treatment with exposure therapy, could be utilized to assess treatment responsiveness in the future. In the literature reviewed above, we identified six studies (Wallach et al., 2009, 2011; Safir et al., 2012; Anderson et al., 2013, 2017; Bouchard et al., 2017) that incorporated behavioral avoidance tasks to probe the behavioral consequences of behavioral interventions. Two studies measured the duration of speech during a public speaking task (Anderson et al., 2013; Kampmann et al., 2016b) as a measure of behavioral change. Interestingly, both of studies were conducted on a population with SAD, and speech duration was not measured in any of the studies involving PSA populations, despite its potential of being an even more realistic behavioral measure for this condition. In conclusion, even though self- and clinician-rated instruments are standard in clinical care, due to their subjectivity, additional measures such as physiological recordings and behavioral assessments are promising, but are used only sporadically. Increased integration of these measures could lead to a more comprehensive and objective evaluation of treatment response.

Most studies have employed treatment protocols spanning several sessions of VRET in line with current treatment guidelines. A recent study examining different numbers of VRET sessions and their impact on treatment outcomes observed that similar efficacy could be achieved with nine or ten sessions, and that even shorter interventions of five or six sessions could result in stable clinical improvement. Expanding the number of sessions beyond ten did not appear to lead to additionally improved therapeutic effects (Jeong et al., 2021). This is in line with a review of n = 28 studies (Krzystanek et al., 2021) recommending eight to 12 weekly sessions of at least 15 min of VRET for the treatment of SAD. Although recent meta-analyses have found no significant effect of the number of treatment sessions to influence the efficacy of VRET against control conditions or in vivo exposure (Carl et al., 2019; Horigome et al., 2020), Horigome et al. (2020) observed a trend for effect sizes to increase with session number, and, in a similar vein, Opris et al. (2012) reported a significant moderating effect of the number of treatments sessions to moderate effect sizes.

While few studies, mostly restricted to the phenotype of PSA, have investigated single-session VRET, significant symptom reductions have been reported, but studies were mainly conducted in small samples and sub-clinical populations. This calls for the systematic evaluation of the potential benefit of single-session VRET in larger, sufficiently powered samples following a randomized-controlled approach, also extending to the phenotype of clinical SAD, in future studies.

4.2. Attrition

Meta-analytic evidence points to relatively low drop-out rates from VRET with an average of 16% (Benbow and Anderson, 2019), which are comparable to drop-out rates from CBT with in vivo exposure (van Loenen et al., 2022). However, reasons for discontinuing treatment appear to differ between virtual and in vivo exposure: While a frequently reported reason for drop-out from exposure in vivo is fear of exposure to the aversive stimulus itself (Benbow and Anderson, 2019), thus constituting a dysfunctional avoidance mechanism that may further aggravate the fear and hinder clinical improvement (Taylor et al., 2012), the most commonly reported reason for drop-out from VRET is failed immersion in the VR environment (Benbow and Anderson, 2019). Thus, in the future, given technological advancement of VR presentation modalities and equipment, such drop-outs from VRET could be reduced by promoting immersion and sense of presence in the VR environment.

While dropout is similar, compliance may be higher in VRET than in standard CBT (Liu et al., 2022). Surveys assessing treatment preferences point to a majority of patients favoring VRET over in vivo exposure therapy (Garcia-Palacios et al., 2001, 2007). This preference may thus increase treatment adherence, given that individuals are more likely to remain in treatment when they have a say in the choice of treatment (Swift et al., 2011). Additionally, offering VRET as a treatment option has been associated with lower refusal rates compared to exposure in vivo (Garcia-Palacios et al., 2007).

4.3. Accessibility

As detailed above, the efficacy of VRET for SAD and/or PSA make it a promising and appealing option for both patients and clinicians in the management of anxiety disorders. Exposure-based interventions addressing social fears can be resource-intensive and time-consuming due to the need to facilitate exposure exercises often outside the safety of a standardized, closely measured and accompanied environment (Reeves et al., 2022; Bouchard et al., 2017). VRET presents a promising alternative and offers several advantages in terms of resource efficiency and accessibility. With VRET, patients have the opportunity to engage in social situations within the safety and convenience of a therapist's office, requiring only the necessary VR equipment. This significantly reduces the resource investment needed for exposure-based treatments, including time and finances (Hartanto et al., 2014; Carl et al., 2019; Emmelkamp, 2005; Miloff et al., 2016; Reeves et al., 2022; Boeldt et al., 2019; Robillard et al., 2011).

Traditional exposure exercises may involve arranging and conducting real-life scenarios, which can be logistically challenging and costly. VRET provides a controlled and standardized environment for exposure, allowing the therapist to carefully tailor the exposure experiences to the individual patient's needs and gradually increase the difficulty level at a pace that is manageable for the patient. Additionally, VRET offers enhanced confidentiality, as patients can engage in exposure exercises within the privacy of the therapist's office. Future developments and technological advancements in consumer-ready VR hardware would even allow clinicians to equip patients with light-weight VR equipment, making it possible to conduct self-guided VRET outside regular therapy sessions as a homework assignment, which is crucial to anxiety and phobia treatment (Cooper et al., 2017; Heimberg et al., 1985; Huppert et al., 2006). Furthermore, access to psychotherapy is often hindered by e.g. stigmatization, physical distance, or long waiting times (Hong et al., 2021; World Health Organization, 2020; Backhaus et al., 2012; Simpson and Reid, 2014). With self-conducted stand-alone VRET implementations, potentially in combination with internet-based therapy modules or online therapist contact, treatment could be delivered to patients with restricted access to care. This way, patients could benefit from exposure therapy without the need to travel to various real-world settings, especially for individuals living in remote areas or those with limited mobility. If developed and conducted along evidence-based principles, self-guided VRET could also be applied within a stepped-care framework in clinical settings. However, self-guided VRET formats are still a novelty (Anderson and Molloy, 2020; Chard and van Zalk, 2022), and their ramification as stand-alone interventions has to be considered very carefully.

4.4. Therapist- vs. self-guided VRET

While the majority of studies have combined VRET with psychoeducative and cognitive elements, and treatment was administered by trained psychologists or psychotherapists, several studies have probed the use of self-administered VRET without the guidance of a therapist. For instance, applying a smartphone-based app using 360° video recordings of a lecture hall in a home-setting without a therapist present, n = 35 participants with PSA took part in three sessions of self-administered VRET following a gradual exposure approach of speaking in front of no audience or an audience of increasing size. After VRET, a significant reduction in PSA symptoms was observed, particularly in participants with high pre-treatment levels of PSA (Stupar-Rutenfrans et al., 2017). In another study, participants with high PSA participated in two weekly self-guided VRET sessions of a public speaking scenario, which resulted in a reduction of anxiety, fear of negative evaluation, and heart rate, with effects remaining for at least one month (Premkumar et al., 2021). In a study in SAD patients who underwent four or more self-guided sessions of VRET comprising either a job interview or a dinner party scenario, SAD symptom reductions were reported post-treatment and maintained after six months (Zainal et al., 2021). In two controlled clinical trials investigating the effects of eight sessions of isolated self-training VRET without further CBT elements over two weeks in patients with SAD, significant decreases in self-reported SAD symptoms were reported in comparison to healthy controls (Kim et al., 2017) and a waitlist control group (Kim et al., 2022).

However, research directly comparing the efficacy of therapist-vs. self-guided VRET using the same setup and protocol for the treatment of PSA or SAD is scarce. Stefaniak et al. (2022) found self-guided VRET to be inferior to CBT in reducing social anxiety symptoms post-treatment and reported significantly higher drop-out rates in self-guided VRET comparted to therapist-led VRET. In a similar vein, in a subclinical population with PSA, a therapist-led one-session VRET intervention complemented by a four-week online program including in vivo exposure also resulted in greater symptom improvements than a self-guided intervention complemented by the same online program (Lindner et al., 2019). Therefore, while self-guided interventions are certainly intriguing given long wait times, accessibility and cost-effectiveness as outlined above, future studies are urgently needed to address the comparability between therapist- and self-guided VR interventions.

Taking self-administered VRET without the guidance of a therapist to a naturalistic setting is also not without its pitfalls: With a therapist present, the selection and intensity of the presented scenarios can be carefully curated and controlled throughout the entire exposure session. The therapist can monitor the patient's condition and behavior, adjust the VR environment in real-time and ensure that the exposure session is conducted in an effective and goal-oriented way (e.g., the patient not referring back to avoidant behavior or attention patterns). Additionally, the presence of a therapist ensures a certain level of safety during VRET, as they can professionally provide immediate care in cases of symptom worsening, severe simulation sickness, personal crises, or the patient feeling overwhelmed or helpless. Finally, self-guided VRET tools or apps call for strict quality and safety controls to ensure that they are construed in line with evidence-based guidelines. While not specific to VRET, a recent meta-analysis has shown that mental health apps oftentimes include techniques for which there is no clear evidence (Lagan et al., 2021) and may even contain harmful content (Larsen et al., 2016).

4.5. Therapeutic alliance

Considering self-guided VRET and other self-guided digital mental health interventions naturally raises the question of the role of the relationship between patient and therapist. As a widely known process variable and predictor of therapeutic outcome (e.g., Fluckiger et al., 2018; Horvath et al., 2011; Martin et al., 2000), therapeutic alliance constitutes an important factor in psychotherapy that should be taken into account when evaluating novel therapeutic techniques and settings, especially considering its role for the efficacy and attrition rates in SAD treatment (Haug et al., 2016). Regarding the therapeutic alliance between therapist and patient, meta-analyses indicate no impairment of the therapeutic alliance by digital mental health interventions of various types (Berger, 2017; Seuling et al., 2023; Tremain et al., 2020) and VRET specifically (Horigome et al., 2020; Meyerbroker and Emmelkamp, 2010). Directly compared, VRET seems to allow for the development of a therapeutic alliance of comparable strength to exposure in vivo (Ngai, 2012; Draheim and Anderson, 2019). So far, in two trials, large correlations were shown between symptom reduction and therapeutic alliance (Ngai, 2012; Moldovan and David, 2014). By contrast, another study (Draheim and Anderson, 2019) did not find evidence for a direct link between therapeutic alliance and response rates in VRET. Thus, the specific role of therapeutic alliance for treatment success in VRET remains to be elucidated in future studies.

The dynamics of therapeutic alliance may differ when comparing therapist-guided to self-guided VRET. The absence of direct guidance and support through a trained therapist during exposure scenarios might raise questions about the impact on therapeutic alliance and, therefore, treatment outcomes. Understanding the interplay between therapeutic alliance and the mode of exposure delivery (be it through VR vs. in vivo or therapist-led vs. self-led) is essential for optimizing the treatment of anxiety disorders. Thus, further research into the role of therapeutic alliance in diverse treatment modalities is vital.

4.6. Variation of VR situations

Research on VRET for SAD and PSA prominently revolves around interventions where patients are confronted with scenarios involving public speaking (i.e., speaking in front of an audience). Public speaking is among the most common social challenges feared in SAD (Stein et al., 1996; Furmark et al., 2000) and has thus become a focal point for VRET interventions. The emphasis on public speaking scenarios in VRET research can furthermore be attributed to the technical and logistical challenges implied by programming VR situations for other situations relevant in SAD, as they may require greater variability and complexity regarding the avatars’ behavior and interactions. Finally, public speaking situations can offer a higher degree of standardization during the exposure compared to other social situations. Still, as with in vivo exposure, providing a variety of exposure contexts would allow for greater generalization across situations in generalized social phobia. Several of the studies reviewed above have offered a broader selection of scenarios for VRET. Apart from public speaking situations, job interviews comprised a considerable amount of exposure situations (e.g., Bouchard et al., 2017; Stefaniak et al., 2022; Kampmann et al., 2016b). Other scenarios include various social challenges at a party (Gebara et al., 2016), having a blind date (Kampmann et al., 2016b), participating in small talk (Klinger et al., 2005), or buying and returning clothes (Kampmann et al., 2016b). Where the tasks exceeded public speaking situations, cafés (Klinger et al., 2005; Geraets et al., 2019), public transport (Stefaniak et al., 2022; Geraets et al., 2019), stores (Bouchard et al., 2017; Klinger et al., 2005; Geraets et al., 2019), or hallways (Bouchard et al., 2017; Klinger et al., 2005) served as locations in VR for conducting the exposures. Still, public speaking situations represent the overwhelming majority of VRET scenarios, thereby representing only a small fraction of the range of potential trigger stimuli for social anxiety. In order to comprehensively address the versatile challenges that individuals with SAD encounter in social situations, it is paramount to more frequently implement VRET scenarios that go beyond public speaking. While the focus on public speaking has provided valuable insights, future research should strive for a more nuanced representation of social scenarios to enhance the effectiveness and applicability of VRET in treating SAD.

Another opportunity for variation of the VR situation is the graduation of the intensity of the situation, as e.g. through increasing or decreasing the number of avatars present during a VR scenario (Reeves et al., 2021). However, the systematic assessment of such graduation in exposure difficulty is relatively uncommon in trials implementing VRET for SAD or PSA thus far, underscoring the need for its exploration in future research endeavors.

4.7. Promises and pitfalls

The continuous advancement in VR technology leads to a growing usability and accessibility of the equipment, fostering gamification and consumer-oriented development of applications, and allows for a more scalable placement of the respective products on the free market. Big global players, like Apple or Meta, have long identified VR as a new modality to present digital content in the future, and are already well-advanced in the process of developing cost-efficient and easily accessible VR hardware and software. Despite these advances, VR still has some limitations compared to what is possible in the real world that are also directly applicable to the use of VR in therapeutic settings. For instance, improved graphical representations aimed at introducing a greater degree of realism in virtual worlds may result in a counterproductive effect – termed the uncanny valley effect – and appear unsettling to the observer, thus making them appear less realistic and engaging. Different approaches have been explored to address this effect, for instance utilizing video recordings of real audiences within VR instead of computer-rendered virtual humans (cf. Lindner et al., 2019). By incorporating real human actors or recorded audience members into the VR environment, the virtual social situations can become more authentic and relatable for patients undergoing exposure therapy. Furthermore, common occurrences such as motion sickness or drowsiness limit the time that can be spent within the virtual environment and sometimes even preclude participation in VR settings altogether. Symptoms most likely arise due to a sensory mismatch between the visual (and sometimes auditory) perception of movement, and the vestibular and proprioception of remaining at one place (Bos et al., 2008; Davis et al., 2015; Howarth and Costello, 1997; LaViola, 2000). However, the advancement in VR presentation technologies is already making progresses towards decreasing, and, eventually, potentially even eliminating simulator sickness (Parsons, 2021).

Still, with VR technology becoming both more advanced and affordable, implications for treatment use are highly promising, making possible a more realistic simulation of even complex situations, and prospectively enabling the inclusion of biomarker-assisted outreads into the treatment process through precise programming and behavioral tracking, as well as making VR technology more accessible also in clinical routine settings. Given the popularity of VR games, in part made possible by VR gear having become smaller, lighter, and less expensive, an increasing number of the population has become familiar with the use of VR for recreational purposes. Future research will have to address the role of previous VR experience and whether it affects VRET adherence and outcomes. In a similar vein, patients’ attitudes and expectations towards VR may also affect openness to VRET as well as treatment outcomes, but have only been investigated sparingly. Patient expectations regarding VRET have been shown to predict clinical change after VRET in SAD (Moldovan and David, 2014; Price and Anderson, 2012), but by contrast, patient expectancies did not predict symptom reduction in a study in PSA (Scheveneels et al., 2019).

Acceptance of VRET appears to be high in general, surpassing acceptance rates of in vivo exposure (Garcia-Palacios et al., 2007), but it remains to be elucidated in future studies whether openness to VR-assisted treatment varies by familiarity with VR, attitudes towards technology, or age group, with younger populations, often termed ‘digital natives’, more frequently engaging with novel technologies and virtual worlds. On a related note, while most studies on VRET were conducted in adults, some studies have specifically adapted and successfully applied VRET for adolescents with anxiety (Kahlon et al., 2019; Sülter et al., 2022). However, to date, no study has systematically adapted and evaluated VRET in elderly populations.

In an effort to increase treatment response rates, recent research has begun to address the potential augmentation of VR interventions, for instance by modifying attentional biases. Bias modification has been shown to be effective for treating anxiety disorders in general (Kampmann et al., 2016a), even though it appears to be less useful for clinical SAD (Heeren et al., 2015). While changing attentional bias in a VR attention bias modification task has not proven successful (Ma et al., 2019), in a recent trial, visual attention to a virtual audience was shown to change as an effect of complementing VRET with an attention guidance training (Rubin et al., 2022). However, at present, even though changes in attention focus may constitute a highly relevant process mechanism of therapeutic improvement during and after exposure therapy (Gregory and Peters, 2017; Kampmann et al., 2019; Norton and Abbott, 2016; Wong et al., 2017; Hofmann, 2000), amplifying standard VRET with interventions aimed at modifying dysfunctional attentional biases in SAD does not appear to result in additional clinical change and provides no incremental advantage to VRET, which is in line with the growing criticism towards attention bias modification (Kampmann et al., 2016a; Emmelkamp et al., 2020; Emmelkamp, 2012).

VRET has also been augmented pharmacologically by administering D-cycloserine with VRET for PTSD patients. This combination has been proven effective in reducing PTSD symptoms (Rothbaum et al., 2014; Difede et al., 2014), with the supplementation of D-cycloserine increasing treatment efficacy over standard VRET in one trial (Difede et al., 2014). For SAD, early studies on D-cycloserine administration during exposure in vivo have yielded promising results (Guastella et al., 2008; Hofmann et al., 2006; Rodrigues et al., 2014), although null effects have also been reported (Hofmann et al., 2013). A recent meta-analysis (Burkner et al., 2017) concluded a small to moderate effect of D-cycloserine administration for SAD patients in particular. Thus, even though studies on D-cycloserine augmentation of VRET for SAD are presently lacking, its application in future VR trials in SAD appears to be promising.

Besides studies in clinical SAD, VRET has also been shown to alleviate symptoms of social anxiety in subclinical, but socially anxious probands (Morina et al., 2015; Wechsler et al., 2021) who may be at increased risk for developing SAD. Thus, VRET may represent an efficacious, cost-effective tool also in the prevention of SAD when applied to populations at risk before fully meeting diagnostic criteria. VR interventions specifically tailored to the prevention of SAD and other anxiety disorders would complement presently available programs (e.g., Norr et al., 2020); Schiele et al., 2021; Schmidt et al., 2017; for review see Domschke et al., 2021; Lau and Rapee, 2011) in an effort to reduce the burden associated with anxiety disorders.

Next to its clinical efficacy, VR also provides considerable advantages for research purposes. Firstly, it eliminates many sources of treatment variation, which are often subtle and therefore difficult to detect or to control. Real-world exposure exercises are by nature subject to many uncontrollable and unstandardized influences, limiting direct comparability from a research perspective. Providing social situations in VR enables researchers to precisely record, administer, and control the majority of these variables, making it possible to not only eliminate confounders, but to also investigate the effects of specific variables on treatment outcomes. This way, the strict standardization of exposures across patients allows to more closely examine the core mechanisms that drive therapeutic change in exposure therapy, to identify key elements of the exposure experience, and to correlate them with behavioral or physiological data at an immense temporal resolution that would be impossible in most in vivo settings (cf. Kroczek et al., 2020). For example, several studies have investigated physiological measures such as heart rate during VRET (Lister et al., 2010; Harris et al., 2002; Kim et al., 2017; Wechsler et al., 2021; Hartanto et al., 2016; van Dis et al., 2021) and for instance reported changes in heart rate along with symptom improvements after treatment of PSA with four (Harris et al., 2002) or ten sessions of VRET for SAD (Hartanto et al., 2016). While successfully treating PSA with VRET, increased heart rate and skin conductance have also been observed during the exposure itself (Lister et al., 2010). These measures have furthermore been shown to decrease over the course of the exposure in high socially anxious participants (Wechsler et al., 2021). By contrast, in SAD patients, heart rate has been observed to decrease during VRET alongside subjective measures of stress and arousal (van Dis et al., 2021). Real-time measurement of physiological data acquired during VRET can even be used to automatically classify the level of anxiety through blood pressure or galvanic skin response (Šalkevicius et al., 2019).

Even though VR, and especially VRET, provides a promising research framework, to this day, evidence on its usage is mostly restricted to small-scale studies on the clinical efficacy for symptom reduction. While providing a highly valuable background for the general evaluation of VRET, these studies are usually conducted in strictly controlled research settings and exclude some factors that are crucial for the translation to clinical practice, such as medication or comorbidities. Next to the documented efficacy of VRET, future research endeavors should aim to broaden the scope of VRET trials by investigating its effectiveness in more naturalistic settings and with larger sample sizes.

4.8. VR-EThics

As with all novel technologies and digital applications, strict adherence to ethical principles is indispensable, especially in the field of mental health care (Marloth et al., 2020). Equally to psychotherapeutic interventions in general, adverse effects of therapy such as treatment failure or symptom deterioration can occur (cf. Linden, 2013). In addition, VRET can provoke simulator sickness or cybersickness (Davis et al., 2014), which describes feelings of nausea, asthenopia, headache, or fatigue commonly occurring during or after extended and/or unfamiliar immersion in VR (LaViola, 2000; Davis et al., 2014) that may negatively impact the VR experience, and therefore possibly VRET outcomes. The increasingly high amount of realism possible in VR settings could lead to sensory overload (or “information overload”; Behr et al., 2005) in vulnerable individuals, and superrealism can pose challenging situations during VRET. Thus, it is of utmost importance that therapists carefully observe patients’ condition and reaction towards the VR environment and guide them safely through VRET sessions. This requires detailed education about possible advantageous and adverse effects of VRET in particular within the process of obtaining informed consent to participate in VRET (Parsons, 2021; Behr et al., 2005). One particular concern also pertains to ensuring the safety and protection of individual patient data when using VR applications. Given that VR technology allows for the electronic collection of vast amounts of data, caution regarding the specifics of the hardware and software is warranted, especially in the context of third-party applications, on-line data collection, data storage and protection against potential manipulation (Marloth et al., 2020; Yellowlees et al., 2012). Crucially, in the case of stand-alone self-led VRET applications, it has to be considered that, as long as they are available through common distribution channels such as app stores, patients using these applications may lack the means to verify if the software is empirically validated and suitable and safe for their specific needs. This is especially critical when considering that newly developed treatment options often raise disproportionally strong expectations in patients, an effect commonly referred to as “therapeutic misconceptions” (Marloth et al., 2020).

5. Conclusions

With its benefits of standardization, accessibility, and patient acceptance, VRET emerges as a valuable therapeutic tool to treat SAD and PSA and offers an alternative and comparably effective approach to traditional exposure therapy. Utilizing immersive technologies, VRET allows patients to virtually confront fear-inducing situations, yielding significant reductions in anxiety symptoms. In addition to its efficacy, VRET represents an excellent research framework allowing for a more standardized, highly experimentally manipulable and controllable investigation of the mechanisms of actions underlying exposure-based interventions. Moreover, the accessibility and cost-effectiveness of VRET make it a viable option for patients with restricted access to care. Self-guided VRET offers a potentially scalable and convenient treatment option, but it requires careful consideration and safety controls to ensure evidence-based practice. Ethical considerations regarding protection of patient data and administration of safe and validated VRET applications should also be prioritized to protect patients. In conclusion, VRET promises to be an effective and accessible treatment option for individuals with SAD or PSA, also opening up novel avenues for their prevention and management.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Conflict of interest

All authors declare no conflict of interest.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Handling Editor: Prof. A. Meyer-Lindenberg