Virtual reality exposure therapy for fear of spiders: an open trial and feasibility study of a new treatment for arachnophobia

Abstract

Purpose

This analogue pilot study examined the feasibility (i.e. preliminary results, safety, acceptability) of a new single-session treatment for adults with a fear of spiders.

Materials

It used state-of-the-art consumer available VR-hardware for therapist-assisted exposure (VRET–AP). The VRET-AP is largely adapted from Öst’s one-session treatment for arachnophobia (Öst, 1987), with the aim of addressing shortcomings of previous VRET treatments, such as marked differences in procedures compared to available and evidence based in-vivo treatments.

Method

Participants (N = 12) were screened for fear of spiders using the Spider Phobia Questionnaire (SPQ), Fear Questionnaire (FQ) and the Behavioral Approach Test (BAT), prior to and directly after treatment in a repeated measures quasi-experimental design. In addition, acceptance and completion rates were measured and participants were interviewed about their experience of the treatment. Mean ratings as well as Reliable Change Index (RCI) for individual trajectories were analyzed.

Results

The results from the preliminary data indicates potential for improvements with large effect sizes (d = 0.90-1,384) in all measurements of spider fear at post-treatment. Reliable Change Index (RCI) analysis showed that spider fear diminished in all twelve participants although the change was certain in only two. None deteriorated. All that responded accepted the treatment and all 11 participants completed all levels in the treatment. No concerns or adverse effects were reported in the interviews which largely confirm the quantitative results.

Conclusion

VRET-AP is a feasible alternative for delivering effective treatment for fear of spiders and the results motivate larger, randomized trials of VRET-AP involving participants diagnosed with arachnophobia.

Keywords:

• Exposure therapy
• one-session treatment
• spider phobia
• arachnophobia
• virtual reality
• anxiety
• CBT

Introduction

Arachnophobia (AP), otherwise known as spider phobia, is defined as excessive fear and avoidance in response to thoughts about or stimuli reminiscent of spiders [1]. When the fear and avoidance cause significant distress to the individual or marked impairment in functioning, it qualifies as a psychiatric disorder (Specific Phobia – Animal Type) according to both the 5th edition of the Diagnostic and Statistical Manual of Mental Disorders [2] and the 11th edition of the International Classification of Diseases [3]. AP is one of the most prevalent forms of specific phobia with a lifetime prevalence of approximately 5%, occurring somewhat more frequently in women than men [2,4,5] Specific phobias, including AP, often develop quite early in life and their presence is associated with an increased risk of developing other anxiety disorders and depression across the lifespan [6].

The gold standard for treatment for specific phobias is in-person cognitive behavioral therapy (CBT), involving gradual in vivo and imaginal exposure to the feared stimuli/situation, including [7] one-session in-vivo treatment [8]. However, most patients with specific phobias, including AP, do not seek treatment for this condition [9], possibly owing to fears about what treatment will involve [10]. Another potential barrier to care is that many mental health professionals are reluctant to use exposure-based therapies because of concerns about the tolerability of the treatment [11]. These concerns are only partly supported by meta-analyses that find that approximately 20% of adults dropout before completing a course of anxiety-focused CBT [12,13].

Virtual reality exposure therapy (VRET) involves standard therapeutic exposure interventions delivered within a computerized virtual environment using virtual feared stimuli experienced in a multisensory way through stereoscopic head-tracker displays, hand/body tracking and binaural sound [14]. Research show that treating specific phobia in VRET can be equally effective as standard in-vivo exposure therapy [15,16], with a recent meta-analysis suggesting marginally lower drop-out rates in VRET than standard CBT (16% vs 19.6%; 12]. However, there is fairly large variance in treatment outcomes between different VRET studies, variance that may reflect marked differences in the virtual environments of the VRET interventions, how the interventions are delivered (e.g. therapist assisted, automated, gamified or not, with/without haptic feedback), and to whom [17–21]. In previous VRET-applications attrition has mainly been attributed to lack of immersion and cyber sickness resulting from the VR applications used [12]. It is widely recognized that further research is needed in order to better understand the necessary and therapeutically active components of VRET and to improve the format to establish their feasibility, acceptability, and efficacy [21]. The primary aim of the present pilot study was to investigate the acceptability and feasibility of a newly designed VR-assisted and therapist led exposure therapy for arachnophobia (VRET-AP), largely based on Öst’s one-session phobia treatment [7], which we believe better reflects the procedures and techniques used in evidence based golden standard treatments than previous VRET applications briefly discussed above.

Based on previous research [12,22], we anticipated: 1) that preliminary results would indicate reductions in the severity of fear as indexed by the self-report measures and behavioral approach in relation to a live spider; 2) that the majority of the participants offered VRET-AP would accept the treatment; 3) that at least 80% would complete the treatment; and 4) would report relatively high levels of treatment satisfaction based on post-treatment qualitative interviews.

Materials and method

Aims, design and hypotheses

The present study was an uncontrolled analogue pilot designed to test out our VRET-A treatment protocol and materials. Furthermore, we aimed to collect preliminary data on treatment effects of VRET-A on fear of spiders. The secondary aim was to investigate the acceptability and feasibility of the treatment in adults highly fearful of spiders and to learn how the treatment was experienced. Preliminary data consisted of participants’ scores on the Spider Phobia Questionnaire (SPQ), Fear Questionnaire (FQ), and a Behavioral Approach Test (BAT). A baseline phase lasting at least five full days was followed by treatment, lasting from 2-4 h, and a 1-2 h follow-up one week after completing treatment. Feasibility and acceptability were assessed by the proportion of patients who accepted the treatment (VRET-AP), treatment completion rate, and via post-treatment interviews involving thematic analysis. All participants gave written informed consent. The study was approved by the regional ethics committee board (dnr 2020-04785). No financial compensation for participation was offered. The VRET-AP was developed by the authors in collaboration with dr. Jonas Ahlstedt at Lund University Bioimaging Center. The authors declare that they have no other competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

The intervention

VRET-AP is largely based on Öst [7] one-session in vivo treatment manual. The application is unfunded at the time of writing and not intended for commercial use. The intervention was delivered via an Oculus Quest 2 virtual reality device that was cable-connected to a gaming computer running Unreal Engine 4. The therapist-assisted exposure to spider stimuli in the VR program was divided into six levels (0–5). Before each level, the participant was informed what the level entailed and the goal of that level. The first preparatory level was nicknamed ‘level zero’ and served as a way to get the participant familiar with the VR equipment and how to move around in the VR environment. It featured free interaction with a small, lockable, red plastic container of the type used to store household food, and small plastic shovel on a table. These objects were chosen since they were used in later levels. Prior to commencing Level 0, the participants were informed that the spider was not currently present, and during later levels, the spider would always be located in the kitchen.

Level 1 started with the VR-spider inside the locked plastic container and the participant in a hallway outside the kitchen approaching the container, moving the container around the table, picking it up and putting it down. When anxiety subdued, the participant was asked which actions, such as suddenly dropping the container or placing one hand under the container whilst holding it, caused more discomfort, and then proceeded to test that out by repeating these actions. In Level 2, the participant is told that the spider is next to and outside the plastic container on the table in the kitchen. In Level 3, the spider was uncontained on the table next to the shovel, with the latter being available to move the spider. When discomfort subdued, the participant was asked to move the spider in ways that would cause more discomfort, such as moving the spider closer, picking it up with the shovel or suddenly dropping it on the floor. In level 4, the uncontained spider sitting on the table was picked up using their virtual hands, visible in the participants’ field of view and acting as their own hand (closely mimicking their movement). Level 5 consisted of the spider being unconstrained in the same way as level four. The participant was asked to lift the spider and throw it straight up over their head. After this no longer led to high feelings of discomfort, the participant was asked to repeat the action with their eyes closed. VRET-AP lasted for between two and four hours depending on the rate of progress through the stages.

VRET-AP therapists

The two therapists were final-year students on the psychology and medical programs, respectively, and both had some clinical experience as well as formal clinical training, although not in VRET-AP. In order to enhance treatment adherence, the treatment protocol was presented and discussed during a two-hour seminar led by one of the authors (JK), a licensed psychotherapist with extensive experience of treating patients with AP with traditional CBT as well as training therapists to provide the same. This was followed by a 4-h seminary involving role-plays of assessments, psychoeducation and the different treatment steps in accordance with the treatment protocol, with continuous feedback from the same author. Thereafter the therapists, under continuous supervision, completed one pilot-session of VRET-AP with a participant with severe AP and who was not included in the study. This pilot treatment case was discussed in depth with author/supervisor where the inclusion and performance of each step in the treatment was assessed by the supervisor and trained until it reflected the actions and approached described in the protocol. The technical aspect of the session was overseen by the VR-programmer ahead of the pilot-session. This included detailed tutoring of the therapists on how to handle the technical aspects, such as set up of the computer and headset, assisting in case of any technical issues as well as regular level switching. All in all, the therapists received three days of training. During the exposure sessions one of two therapists, randomly assigned, acted as the therapist.

Participants

Participants were recruited to the study by the use of flyers or by a posting on social media. Thus, participants initiated contact by e-mail and were then randomly selected to be treated by one of the two therapists. At this point, they were also pseudonymized in order to protect their anonymity. Criteria for exclusion were physical inability to perform VRET-AP (i.e. blindness) and self-reported comorbid psychiatric illness (asked for during recruitment). Criteria for inclusion were self-admitted fear of spiders or arachnophobia which in addition was was measured as described below. All participants gave their informed and written consent to participate in the study.

19 people expressed an interest in participating in the study. One met the exclusion criteria, and six, prior to filling out any measures, stopped responding to emails for reasons unknown. Thus, twelve participants were included and underwent the VRET-AP. Three participants could not be interviewed due to their time constraints, resulting in a total of nine participant interviews. Consort flow chart is presented below (Figure 1).

Figure 1. Flow diagram of the progress through the phases of a single group repeated measurements trial.

The participants were all women ranging in age between 18 and 54 (M = 28.4).

Measurements

SPQ consists of 31 statements (true/false) concerning spider fear, resulting in total score ranging from 0 (no fear) to 31 (extreme fear). According to a Swedish study, people diagnosed with arachnophobia showed mean ratings at 23.76 (SD = 3.80) and controls at 3.80 (SD = 5.02). The test-retest reliability is r = 0.94 (three weeks) and r = 0.87 (twelve months) and Chronbachs alpha is .90 [23].

FQ consists of 24 questions measuring a range of phobias, including arachnophobia. The first 17 items are rated from zero (‘would not avoid it’) to eight (‘always avoid it’), resulting in a total from 0 (never avoid) to 136 (always avoid). The participants also rate phobic symptoms from zero (no phobias present) to eight (very severely disturbing/disabling phobic symptoms). Finally, the questionnaire include a subsection (with 6 questions) where the participants rate their level of general anxiety and/or depression scoring how troubled they are from zero (hardly at all) to eight (very severely troublesome).

BAT was used to measure the ability of a participant to stepwise approach a living spider. The spider was placed visible in a glass jar on the opposite side of the ten feet wide and thirteen feet long room. The participant was asked to ‘approach the spider as far as they can’ and that they could opt-out whenever they wanted to. If being able to get within reach of the jar, the participant was asked to touch it, and then to pick it up. After this, they were asked to remove the lid of the jar. If this was done, they were asked to touch the spider with a finger. If successful, they were asked to pick up the spider and hold it for at least 20 s in the palm of their hand. Depending on which step they were on when they aborted the test, a different score between 0 (refusing to enter the room) and 12 (holding the spider in their hand for twenty seconds or more) was given. A higher score indicated a higher capability to interact with spiders, and in effect less fear of spiders.

Qualitative interviews

The post-treatment one-hour phone interview was held seven days after the treatment and recorded. It was held by the therapist that did not perform the treatment. Thus each therapist did five interviews. It was semi-structured with open questions and flexible follow up questions when needed.

The interview covered thoroughly how the VRET-AP was experienced in anticipation of and during the treatment, including questions about the exposure and the therapist as well as the VR-environment (specifically the VR-spider). It asked about what enhanced or disturbed the illusion of reality and some questions were posed on the specifics of the used application, such as progression and level switching, as well as specific questions about the how the spider and virtual environment was presented and what would improve their experience of the treatment. Finally, it covered the week after the treatment and how changes relating to spider fear (if present) was perceived. The participation codes were used to designate from what participant the different citations derives.

Thematic analysis

Interview data was analyzed by one of the therapists (JA) using reflexive thematic analysis [24] and later reviewed by and discussed in depth with the clinical senior researcher (JK). Transcription of each interview was done immediately after it was completed (by JA). After all interviews were completed, the transcriptions were read closely once more. This second reading of each interview was largely data-driven and interpretation was held at a minimum. Statements were coded and when codes were frequently discovered, these were compared to the codes found in the other interviews. By using this approach themes started emerging that were mostly descriptive, and, among other things, frequencies of similar statements or codes was noted. This gave an indication to prevalence in the interviews of certain codes. The analytic process went back and forth between JA and JK After jointly deciding on the final themes.

Procedure

All meetings were held at Lund University Institute of Psychology, except the post-treatment interviews that were held using a video conference-tool (Zoom).

At the first meeting, the participant got written and verbal information on the study and was given a form of consent to fill out prior to attending the second meeting. Then a clinical interview was held, in which the participant was asked to give a thorough description of their fear of spiders, and other psychiatric illnesses or disorders was assessed. If meeting inclusion criteria, measure 1 including SPQ and FQ were filled out.

The second meeting was held five days later, in order to provide ample time to consider their participation. It started with collecting the signed form of consent and giving psychoeducation on the exposure and anxiety, how VRET-AP was to be performed, and how the technology used in the project works. The notion that the therapy is based on voluntarism, and that the participant would not be deliberately surprised or forced to do something during exposure that the participant did not wish to do was emphasized. After this, the participant was asked to repeat what they thought to be the idea behind the therapy, in order to make sure that the information given was received. Following this, the participant once more (measure 2) filled out the SPQ and FQ and did the BAT. This concluded the preparatory work, after which the VRET-AP commenced.

The VRET-AP was divided into levels as described above. During every level, the therapist gave encouragement and avoided reassuring the participant (instead inviting the participant to try their fears out). The participant was repeatedly asked to rate their discomfort from a scale of one to ten in subjective units of discomfort (SUDS), ten being the most uncomfortable, and to observe how and whether it varied during exposure. The goal behind SUDS was to allow the participant to reflect on how they were feeling in order to let them experience how their discomfort slowly vaned. Each action and level described above were repeated until the participant either felt ready to advance to the next level, or their SUDS got markedly lower (usually at three or lower, although participants occasionally wanted to proceed in spite of rating higher SUDS). After level 5 and the VRET-AP was completed, BAT was then repeated in order to see if there had been an increase in ability to approach a real spider. Two participants did not partake in this repeat of BAT due to lack of time and energy respectively after treatment. Lastly, the participant was instructed to maintain the progress they had achieved during the VRET-AP by for example removing spiders in their homes (if they had a household containing more than themselves) and by looking at videos or pictures of spiders.

Meeting three was scheduled a week after the treatment and consisted of filling out SPQ and FQ, as well as a short recap of how the participant had been doing in relation to their fears after treatment. Lastly, the one-hour interview was performed.

Statistical analysis

Within-group paired T-tests were performed in order to determine whether changes in ratings on SPQ, FQ and BAT-progression had occurred. Two-tailed tests were used to compare the means from baseline and pre-treatment to determine whether the ratings were stable over time (i.e. no difference). The one-tailed tests were used when comparing baseline to the dependent measures from post-treatment. All t-tests were done using SPSS 27 and the level of significance was set to p < 0.05. In addition, an intention-to-treat analysis were done for the BAT statistics, using a mixed model regression estimation for the two missing participant measures.

An a priori power analysis was performed for sample size estimation. Based on data from Odgers et al. [25] meta-analysis on the efficacy and efficiency of single session exposure for specific phobia, effect sizes on symptom measures was found to be 1.78 for pre- to post-treatment. Thus, with an alpha = .05 and power = 0.95, the projected sample size needed for an effect size of 1.5 (GPower 3.1) is approximately N = 8 for the pre- to post-treatment analyses in the present pilot.

Within groups Cohen’s d_av effect sizes were calculated using the pre- and post treatment means and the average standard deviation of both repeated measures as a standardizer [26].

To ensure that any difference between the baseline and post-treatment measurements was not caused by measurement error of the outcome instruments and to examine individual variance, sometimes masked by mean differences, a reliable change index (RCI) was calculated. It was based on the Jacobson and Truax model for clinical significance [27]. Reliable change scores were calculated based on the post treatment and the baseline scores for each participant, and the standard error of the difference between these two scores. The change is considered reliable (p < 0.05) if the RCI is greater than 1.96 or below −1.96. Thus, each individual can be classified as (1) improved (RCI greater than 1.96), (2) unchanged (RCI between −1.96 and 1.96), or (3) deteriorated (RCI less than −1.96).

Results

SPQ mean ratings at baseline were 18.30 (SD = 4.03). There were no statistical difference between measure 1 (the first interview) and 2 (the treatment session) on any of the dependent measure means, indicating that no spontaneous improvements occurred in these five days. The baseline statistics for SPQ was t(11) = .920, p = .377, and for FQ, t(11) = −1.773, p = .104. For BAT only pre- and post treatment measures were taken, thus the BAT baseline stability was not examined, The results are displayed in Table 1 below.

Table 1. Means and standard deviation for participants’ ratings on the spider phobia Questionnaire (SPQ, n = 12), fear questionnaire total (FQ, n = 12) and behavioral approach test (BAT, n = 10) at baseline and post-treatment.

	Baseline	Post-treatment	Baseline vs. post-treatment
scale	M (SD)	M (SD)	p	dav
SPQ	18.30 (4.03)	13.80 (2.97)	.002	1.01
FQ	6.50 (.90)	5.25 (1.28)	.009	0.99
BAT	6.40 (2.59)	10.00 (1.25)	.003	1.39

Paired samples repeat measure t-test results and effect sizes for baseline versus post-treatment.

SPQ = Spider Phobia Questionnaire, FQ = Fear Questionnaire, BAT = Behavioral Approach Test, d_av = Cohen’s d_av.

A significant decrease (M = 18.42 to M = 14.25) in SPQ means ratings was found between pre- and post-treatment, t(11) = 3.994; p=.002 with mean change 4.5 (SD = 3.72) and a effect size of d_av = 1.01.

Furthermore, a significant decrease in FQ main phobia mean ratings was found between pre- (M = 6.50) and post-treatment (M = 5.25), t(11) = 3.191, p = .009, with the effect size d_av =0.99. The mean change was 0.5 (SD = 1.18) No change was found for FQ full range, t(11) = 1.593, p = .139.

BAT showed a significant mean increase (improvement) of 3.6 steps (SD = 2.80, range) at post-treatment, t(9) = −3.593, p = .004, d_av = 1.385. Using a fixed-effects-only regression model that accounts for missing data (intent to treat), the significant improvement after treatment was confirmed, t(17.935) = 23.287, p < .001, B = 10.00, 95% CI [9.10, 10.90]. Eight of the ten participants who finished BAT after treatment reached step 10 to 12 of 12 (touching or holding the spider) and all participants were able to remove the lid from the jar.

The RCI scores for each participant are displayed in Table 2 below.

Table 2. Participant reliable change index (RCI) for the dependent measures at post treatment compared to base-line measures.

	SPQ	FQ	BAT
1	1.66	0	1.07
2	0	0
3	0.83	0.92	0
4	1.38	0
5	1.66	0.92	0.72
6	0	1.84	2.15*
7	0	0	1.07
8	0.28	1.84	1.07
9	1.11	0	0.72
10	1.94	2.76*	0.72
11	3.32*	3.68*	3.58*
12	1.66	1.84	1.79

A reliable change less than –1.96 indicates deterioration and greater than 1.96 indicates improvement. SPQ = spider phobia Questionnaire, FQ = fear Questionnaire, BAT = behavioral approach test. *=reliable change evident.

Inspection of RCI-scores in Table 2 shows that no one rated a significant deterioration in any of the dependent measures. All RCI scores were in the direction of improvement, whereof two participants showed significant improvement on FQ and BAT respectively and one improved significantly on FQ.

Completion and acceptability rates

6 out if 18 (33%) that registered interest and were offered the initial clinical interview did not respond for reasons unknown. All that responded accepted the treatment and all but one of the 12 participants who choose to partake completed all levels in the treatment. The participant who opted out did so because of fatigue on grounds of an ongoing pregnancy in combination with the fairly long duration of the treatment.

Interview results

Due to space constrains, the full interview results will be fully published elsewhere and only a brief description with select examples, relevant for evaluating the acceptance and safety of treatment will be presented below.

The thematic analysis of the interviews resulted in three main themes. The first, Fear and Loathing in Virtual Reality, contains disclosures of how fear and ambivalence often were felt while expecting the treatment, but also determination and empowerment. One participant (JA1) told that ‘When I first saw the spider, I had this weird feeling as I realized, okay, this is actually going to be tough. That made me pause for a while, to recalibrate’ - (JA1). During treatment the most common term used when describing how VRET-AP was perceived was ‘exciting’. All participants also mentioned experiencing fear, often surprisingly intense, during exposure. The spider was described as ‘gross’, ‘scariest imaginable’, ‘realistic’, ‘big’ by the participants. Interestingly, no participant reported cybersickness, something we directly asked for. Also, no statements indicated that the VRET-AP was experienced as unacceptable or unsafe, even though high levels of fear and loathing were commonly described. On the contrary many participants wished for increased variation, for example the spider sitting on a bed or in a web in the corner of the room. Several participants, among them JH2 and JA8, said they believed completing the treatment more than once would be even more helpful and two participants said they would never have dared to try out in-vivo exposure therapy prior to VRET-AP (JA9 and JA7), but that they now considered it if they ever needed it later. The remarks about having a therapist present were roughly mixed into two categories, ‘comfort’ or “suspicion’. The last involved fantasies that the therapist would somehow trick them during the treatment or in other ways expose them to real spiders while they were in the virtual environment. Others insisted they would not have been able to perform the tasks on their own.

Furthermore, Several participants said they had seen a spider between the treatment and the interview; ‘I met with spiders on two occasions, and it went better than I had expected. I tried to breathe, and I thought that I had done so well in treatment so why not now’, JA8. Many participants described pride and described increased autonomy. However, even after treatment some safety behaviors and remaining fear of spiders were described, although the fear was easier to handle. Remembering that they were able to hold the VR-spider as well as the BAT-spider before stirred confidence and was used as a coping mechanism. Also, most participants reported not avoiding spiders anymore and some even actively sought out challenges; ‘Now I want to continue and lower my fear. My boyfriend found a spider in our living room, and I picked it up and threw it out of the apartment. I would never have done that before’ (JH2).

In the second theme Task Progression participants described how they challenged themselves beyond the instructions given by the therapist during exposure. The graded exposure and levels of increased difficulty was commonly mentioned as helpful for moving forward through the treatment and achieving the ability to later touch the real spider during BAT (which eight out of ten participants managed); ‘What had the most effect on me was the joy of completing all the goals. There was a triumph to each step passed’ – JH10.

The last and third theme Presence contained disclosures on how the illusion of reality was experienced, enhanced or broken during VRET-A. Several participants reported that they were surprised that they forgot that they were wearing VR-glasses or that the experience was in fact virtual and not real. However, accidental movement in the VR-environment and the VR-glasses having bad fit or sliding down causing blurriness hampered the illusion of reality and progression of exposure. Also, the variation in how the spider moved when standing still quickly became familiar and was described as a reminder that it wasn’t real.

Discussion

In the present pilot study, we piloted a novel therapist led VR exposure treatment, largely similar to Öst’s one-session golden standard for arachnophobia. We tried out the materials involved and collected preliminary treatment effect data using measures of spider fear (SPQ, FQ) and the assessed ability to approach a real spider (BAT). We also investigated the acceptability and safety of the treatment through completion rates and interview data.

Our first hypothesis was that preliminary results would indicate reductions in the severity of fear and increased ability to approach a live spider after the treatment. Although no increase or decrease regarding participants’ symptoms of other phobias was evident in the Fear Questionnaire, the expected decreases in the participants’ fear of spiders were found in both the SPQ as well as in the participants’ main phobia ratings in FQ. Moreover, the BAT a mean improvement was 3.6 steps, meaning that most participants could touch or hold the spider after treatment as opposed to not being able to touch the closed jar before. The observed effect sizes were large, even though great caution has to be taken when inferring from within groups comparisons and the small sample size. The analysis of reliable change on an individual level showed that none of the participants’ spider-fear increased. In fact, the ratings of the dependent measures increased reliably on five occasions. However, most change scores fell between 1.96 and 0. Therefore, although in direction of improvement, the improvements are statistically uncertain according to RCI for most of the participants. The upshot is that the preliminary results support our hypothesis and is reasonably well in line with previous studies on VR treatments of specific phobias [28,29], and also arachnophobia specifically such as Miloff and coworkers series of studies on a gamified exposure [8] Furthermore,

Furthermore, we expected that the majority of the participants offered VRET-AP would accept and complete the treatment, and that the treatment was perceived as safe and helpful. These expectations were all largely supported by the results. VRET-AP was tolerated by eleven of our twelve participants and partaking seems to have had minimal adverse effects. One participant however experienced fatigue at the end of treatment and opted out. The fact that this participant was pregnant at the time may have contributed to the fatigue. All participants (when asked in the later interviews) denied experiencing cybersickness during the treatment, even though player movement was frequent in VRET-AP. A majority of the people who initially signed up for the study accepted the offer of treatment and of those treated all completed all levels during exposure. As discussed above, dropout rates for in in-vivo exposure for specific phobia have been found to be around one in five [12] and in previous VRET around 16% [12]. In comparison in the present results one person out of twelve opted out from treatment. These results may provide some hope that VRET-AP has the potential to reach people suffering from specific phobias, whom otherwise may have opted out or left the treatment prematurely, although the small number of participants in the present pilot warrants further research before any conclusions can be drawn.

One reason found in previous research [12,30] for participants dropping out from treatment in VRET were low immersion and associated lack of fear activation in the participants. Furthermore, Slater and Sanchez-Vives [31] found that for VR to be effective, the user need to feel an illusory sense of reality in the virtual environment. In relation to this, it should be mentioned that the technological aspects of the present VRET-AP, which to some degree was touched upon in the participants’ interview, could be improved further. For example, accidental movement in the VR-environment and the VR-glasses having bad fit or sliding down causing blurriness were mentioned as hampering the illusion of reality and progression of exposure. Also, most participants asked for increased variation in for example how the spider moved when standing still, but also having it running on the floor or on the table in an unpredictable way. Larger variation in where the spider was situated was also mentioned, for example on a bed or sitting in a web in the corner of the room. In graded exposure, variation is often introduced in a linear manner with increasing difficulty (e.g. 7]. However, Craske and coworkers suggests that altering linear progression by introducing shifting of the order of exposure may further increases treatment outcome [32]. Given that Shiban and coworkers [19], were able to significantly increase outcome effect just by changing the colors of the surrounding walls during exposure with virtual spiders, even small increases in variability in the present VRET-AP could have impactful effects, which is an interesting avenue for future research. During our time spent working on the project, some problems with the program used appeared. These problems ranged from being able to walk through walls in certain places, long loading times to that the computer program occasionally crashed shortly after starting it. Along with this, the VR-technology used (Oculus Quest 2) did not work to an acceptable level without using a cable between the headset and the computer, something that proved to be an issue at occasions as participants could step on the cable, ending the program. Finally, the participants’ hands in the VR-environment had a ‘robotic’ look to them, which may reduce immersion. In the future, striving towards greater immersion while in the virtual environment could result in even greater beneficial effects for the participants and in turn further help them overcome their fears.

Even though the results are encouraging, it must be noted that pre- to post effects generally are much larger than in randomized controlled studies [33,34], and therefore the large effect-sizes found here should be interpreted with great caution. They cannot be taken as evidence that the treatment actually leads to reduced fear of spiders. Caution is also warranted by the fact that the participant was exposed to a spider in various degrees in the behavioral approach test before VRET-AP. This exposure may have acted as a confounder.

The current treatment is deliberately designed to be therapist-led. This was done in order to follow golden standard treatment guidelines and to provide opportunities to tailor aspects of the exposure to the participants’ specific beliefs about threats associated to the stimuli during treatment. However, the presence of a therapist (and not just a computer technician) during VRET-AP may introduce effects of non-specific factors (such as the quality of the therapeutic relation). Such factors may enhance the results independently of the intervention, though their impact, if any, are still under debate (for a meta-analysis and discussion of this topic, see 35]. The possible effects of non-specific factors could be mitigated and controlled for in a between groups design in future research.

Some additional limitations are related to our participants. All participants were female. This means our findings applicability to males suffering from fear of spiders is limited. However, since anxiety disorders in general and spider phobia specifically, is more common among females [4] and as other studies on the treatment of arachnophobia report above 80% female participants [18,36] this was expected. Even so, future studies need to establish whether treatment effects are the same regardless of gender.

On a final note, our means of recruiting participants (who volunteered after seeing advertisements) implies a risk for a selection bias towards people with an intrinsic willingness to take an active role in the reduction of their level of phobia and therefore not being entirely representative of the larger population of people suffering from phobic fears. Perhaps most importantly, this is an analogue pilot study, meaning that participants were not formally diagnosed for arachnophobia. That said, our participants were highly fearful of spiders. The BAT measure at pre-treatment resulted in a mean of six steps, meaning that the average participant approached the table with the jar containing a real spider, but could not touch the jar. Moreover, the participants’ mean SPQ rating were 19.00 and thus more comparable to a clinical sample diagnosed with arachnophobia (M = 23.76), than with psychology students acting as a reference group at M = 5.02 [23]. Finally, we did not systematically record the SUDS. In future evaluations of the treatment, SUDS should be systematically recorded and the relation of SUDS-measures and final outcome investigated.

In spite of the many limitations of this pilot study, VRET-AP was found to be promising. Our preliminary outcome data indicate that the treatment may reduce fear of spiders, and we believe that the treatment is safe as no adverse effects or side effects were found, seen or reported in the interviews. We also believe that our preliminary results and the observed improvements motivate carrying out a larger randomized trial of VRET-AP with participants diagnosed with arachnophobia in a clinical setting. Such research could entail comparing VRET-AP to traditional one-session exposure therapy, although the expected difference taking the present and previous research into consideration is fairly small and therefore fairly large sample sizes may be required to detect real differences between the treatments, making such a study both time-consuming and expensive.

Arachnophobia is a common source of anxiety within the general population [6]. As discussed above, this source of anxiety can be successfully treated with the use of Öst’s one-session-therapy, although such traditional exposure therapy (also for other types of phobias) require that the fearful object or situation can be experienced in the presence of therapist [7]. This can prove difficult during for instance the winter months (when spiders are hard to find), if the phobia is something that one cannot be exposed to hygienically (such as a phobia of blood) or if the phobia requires a specific setting (such as a fear of flying). Virtual Reality could serve to ease the availability of treatments for these phobias such as, to reiterate our previous examples, provide a virtual spider during the winter months in a cold country, to show a bleeding virtual person, or to place the phobic in a flying airplane without actually needing a physical airplane. Other computer aided treatments are available such as immersive 3D treatment, used for example by Minns et al. [37]. In comparison, VR et al. lows the participant to interact fully with the environment. This interactivity is advantageous in our opinion, as it enhances flexible exposure and makes the treatment more closely resemble in-vivo exposure treatment.

Conclusion

In conclusion we found that the preliminary results from the main measures indicate that VRET-AP may lead to decreases in participants self-assessed severity of arachnophobia and an increase in their willingness to approach a live spider. Furthermore, the treatment was mostly accepted and experienced as safe. While our findings are carefully optimistic as to the potential of VRET-AP, they are based on a small sample using participants’ own baseline measures as comparisons. Clearly further randomized studies are needed to establish the effectiveness of VRET-AP before the treatment is to be recommended in routine care, and while the availability and reliability of virtual reality equipment have increased markedly in the recent years, VR’s potential in a medical or psychological setting is still a field where much research is left to be done.

Ethics approval statement

The study was scrutinized and approved by the Swedish Ethical Review Authority, DNR 2020-04785.

Ethics and integrity statement

All data is fully anonymous. A licensed psychiatrist and psychologist with experience within the field developed the treatment and oversaw all treatment. The participants had access to further treatment through the researchers after the intervention if needed. The research was scrutinized and approved by the Swedish Ethical Review Authority, DNR 2020-04785.

Patient consent statement

All participants gave their written and informed consent to participate.

Permission to reproduce material from other sources

No material from other sources is reproduced in this research.

Acknowledgements

Dr Jonas Alhstedt at Lund University Bioimaging Center for designing the VR-environment and providing technical assistance during data collection, and professor Sean Perrin at the Department of Psychology at Lund University for reading and providing valuable insight on our manuscript.

Disclosure statement

No potential conflict of interest was reported by the authors.

Data availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Additional information

Notes on contributors

Jacob Andersson

Jacob Andersson, LP, is a Licensed Psychologist in Stockholm child and adolescent psychiatric care. His research interest concern technology’s potential to innovate and improve mental health treatment with a focus on the application of virtual reality for psychotherapy interventions.

Joel Hallin

Joel Hallin, MD, is a Physician at Lund University hospital, with a research interest in technology and its various applications within the field of medicine, ranging from AI to VR.

Anders Tingström

Anders Tingström, MD, PhD, is a Professor and Chief Physician in Psychiatry at Lund University, Sweden. He has centered his research on studies of neuronal plasticity in animal models of depression. This specialized work contributes to the understanding and treatment of depressive disorders.

Jens Knutsson

Jens Knutsson, LP, PhD, is an associate Professor at Lund University, practicing in the field of adolescent bipolar disorder. Knutsson’s research interest is clinical psychology with a focus on treatment effects and processes within psychotherapy of anxiety disorders and multi-family treatment of adolescent bipolar disorder.