Is VRET as Effective as In Vivo Exposure? What Meta-Analyses Show

Why Comparing VRET to In Vivo Exposure Is the Right Clinical Question

For a licensed psychologist weighing whether to bring virtual reality into practice, the first question is not whether VRET works in absolute terms, but whether it is at least as effective as traditional exposure with real stimuli. In vivo exposure has been the reference intervention for specific phobias, social phobia, and panic disorder with agoraphobia for decades, with clinical guidelines such as NICE (CG113, 2011) positioning it as a first-line treatment alongside cognitive-behavioral therapy.

The non-inferiority criterion matters because, if the clinical effects are comparable, VRET brings relevant operational advantages: fine-grained control of the stimulus, reproducible grading across sessions, the ability to expose patients in practice to contexts impossible to recreate on the street (a plane taking off, a large dog off-leash, a panoramic elevator with open space underfoot), and, according to several studies, a lower initial refusal rate from patients.

The literature has taken time to converge on an answer because it requires adequately powered meta-analyses, a reasonable variety of target disorders, and credible active comparators. This article reviews the main meta-analyses that support the current answer, their nuances, and their limitations, avoiding the shortcut of presenting VRET as a silver bullet.

Powers and Emmelkamp (2008): The First Adequately Powered Meta-Analysis

The work by Powers and Emmelkamp (2008), published in the Journal of Anxiety Disorders, was one of the first meta-analyses with enough primary studies to allow formal comparisons between VRET, in vivo exposure, and control conditions. They pooled thirteen controlled trials and reported a large effect size for VRET versus control, and a non-significant difference versus in vivo exposure.

What matters most for the clinician is that the authors showed the difference between modalities, far from clearly favoring the real stimulus, could slightly favor VRET in some sensitivity analyses. That reading has been replicated in later meta-analyses and points to a prudent conclusion: the format of the stimulus matters less than the underlying therapeutic process, provided the virtual scenario evokes presence and conditioned fear sufficiently.

The study's limitations are made explicit in the article itself: a limited number of primary studies, heterogeneity of diagnoses (acrophobia, fear of flying, arachnophobia, social phobia, and panic disorder mixed together), modest sample sizes, and mostly comparators that were not especially strong. It is therefore best read as a robust early signal rather than as definitive evidence.

Opriş et al. (2012): A Replication With a Focus on Clinical Maintenance

The meta-analysis by Opriş, Pintea, García-Palacios, Botella, Szamosközi, and David (2012), published in Depression and Anxiety, replicated the conclusion of the 2008 work with an expanded sample of studies. They reported that VRET is effective versus wait-list and comparable to in vivo exposure, with maintenance of effects at twelve-month follow-up in the studies that assessed that measure.

The involvement of Cristina Botella and Azucena García-Palacios, both from the research line at Universitat Jaume I and Universitat de València, is a notable hallmark: this group has been a pioneer in developing and clinically validating VRET in Spain, with sustained publications since the early 2000s. It is useful for the international reader to know that part of the evidence base supporting VR practice comes from Spanish university clinical research teams, not solely from the English-speaking world.

The emphasis on maintenance is clinically important: an intervention that produces post-treatment improvement but loses effect by six months has limited therapeutic value. The studies included in Opriş et al. point to a reasonable persistence of the effect, in line with what is observed in well-protocolized in vivo exposure.

Carl et al. (2019): The Most Highly Powered Meta-Analysis to Date

Carl, Stein, Levihn-Coon, Pogue, Rothbaum, Emmelkamp, Asmundson, Carlbring, and Powers published, in 2019, the largest meta-analysis available to date. They included thirty randomized clinical trials and approximately one thousand fifty-seven participants, grouped under anxiety and related disorders (specific phobias, social phobia, panic disorder, agoraphobia, and PTSD).

Their results are consistent with the earlier narrative. VRET was significantly superior to control and, versus in vivo exposure, the authors found no evidence of inferiority: differences between modalities were small and, depending on the sensitivity analysis, favored VRET in some comparisons. The question of whether one modality strictly outperforms the other remains open and probably depends on the disorder, the stimulus, and the quality of the intervention.

The authors emphasize two limitations clinicians should internalize. The first is the heterogeneity of VRET protocols across studies: number of sessions, session length, exposure hierarchies, and headset characteristics vary widely, making it difficult to extract a single reference protocol. The second is that blinding patients and therapists is impossible in this type of intervention, which opens the door to demand and expectancy biases that randomization does not control for.

Wechsler, Kümpers, and Mühlberger (2019) and Fodor et al. (2018): Recent Nuances

Wechsler, Kümpers, and Mühlberger (2019) published a systematic review that emphasizes the methodological quality of primary studies and warn that part of the field rests on trials with small sample sizes and suboptimal comparators. Their reading does not contradict Carl et al.'s conclusions on non-inferiority, but it invites more moderate causal claims: VRET appears to be as effective as in vivo exposure across most specific phobias studied; it is not proven with the same robustness as the effect of the exposure mechanism itself.

Fodor, Coteț, Cristea, Parsons, Mosoiu, and Cuijpers (2018) update the picture with a meta-analysis specifically focused on efficacy and mechanisms. They confirm moderate-to-large effect sizes in anxiety disorders and support a reading of clinical equivalence versus CBT with in vivo exposure. An interesting nuance is their mediator analysis: the degree of presence evoked by the virtual scenario emerges as a relevant predictor of symptom change, which connects directly to the technical quality of the environment and to the models of presence and immersion proposed by Slater, Wilbur, and Schubert.

The Valencia Research Line: A Spanish Contribution to the Evidence Base

The team led by Cristina Botella and Rosa Baños at the University of Valencia and Universitat Jaume I is one of the international reference groups in VRET. Their work includes controlled studies on arachnophobia (García-Palacios, Hoffman, Carlin, Furness, and Botella, 2002), fear of flying, claustrophobia, panic disorder with agoraphobia, and complicated grief, as well as scenario development for PTSD.

For a licensed clinical psychologist practicing in Spain, this proximity matters for three reasons. The first is linguistic and cultural: part of the protocols were validated with Spanish-speaking patients and an Iberian clinical sensibility. The second is transferability: the diagnostic criteria, exposure hierarchies, and inclusion/exclusion criteria used are closely aligned with the Madrid Official College of Psychologists (COP-M) and the Spanish public health system. The third is academic continuity: some of today's doctoral candidates and clinicians who are integrating virtual reality into private practice trained directly within this research line.

We cite this tradition not to credit VRET by proximity, but to point out that when a licensed psychologist reviews the literature, they need not rely exclusively on American or English-language studies: the evidence base includes work published by Spanish university clinical research teams, using diagnostic criteria compatible with practice in Madrid or Barcelona.

Methodological Limitations Worth Reading Carefully

No meta-analysis is better than the primary studies that compose it, and in VRET several problems recur. First, protocol heterogeneity: the number and length of sessions, headset type, 3D model quality, intensity of the evoking stimulus, and the therapist's role vary across studies. This means that when we say "VRET is effective," we are talking about a family of interventions, not a single homogeneous one.

Second, the question of blinding. In psychotherapy it is structurally difficult to blind the patient and the therapist. This does not invalidate the findings, but it does require interpreting them as evidence of combined effectiveness (intervention plus expectancy plus alliance) rather than as proof of an isolated mechanism.

Third, the quality of the comparator. When the active group is in vivo exposure well protocolized by trained therapists, the comparison is fair; when it is an impoverished version of the standard treatment, the reported equivalence may overestimate VRET's relative efficacy. Carl et al. (2019) discuss this point honestly.

Fourth, sample sizes: many primary studies work with fewer than sixty participants, which limits the power to detect modest differences between modalities. Finally, generalization: the evidence is robust for specific phobias and reasonable for social phobia and panic disorder, but more limited for complex PTSD, severe agoraphobia with significant comorbidity, or obsessive-compulsive disorder, where studies are fewer and less well powered.

Another limitation worth mentioning is the age of part of the evidence base. The Powers and Emmelkamp work dates from 2008, meaning it captures primary trials conducted with VR technology considerably less advanced than what is available today. Lower-resolution headsets, higher latency, tracking limited to three degrees of freedom, and visually less plausible scenarios characterize part of the foundational evidence. This introduces a conservative bias regarding current efficacy: the observed effects were produced with hardware less effective at evoking presence, and it would be reasonable to expect that with contemporary technology, effect sizes would be at least comparable, probably somewhat larger in better-designed scenarios.

Finally, ecological representativeness deserves consideration. Participants in a clinical trial tend to be patients with a relatively "clean" diagnosis, without much comorbidity, willing to undergo randomization, with schedule availability and motivation to complete the protocol. The real-world patient in private practice may have a rather different profile. Meta-analyses tell us what happens under ideal conditions; clinical practice requires cautious inference about what holds up under real-world conditions.

What This Means for Your Practice

The honest reading of the meta-analyses is that VRET is a clinically equivalent alternative to in vivo exposure for the indications where it has been most studied (specific phobias, social phobia, panic disorder with mild to moderate agoraphobia, PTSD within protocols with specific supervision). It is not absolutely superior to well-delivered traditional exposure, but it is not inferior either, and it brings relevant operational advantages that can tip the balance depending on the clinical case.

In practical terms, the decision to bring VRET into a practice should not be framed as "VRET instead of in vivo exposure" but as "VRET added to the existing repertoire." For specific phobias with triggers that are difficult to reproduce in practice (dog phobia, fear of flying, acrophobia, elevator claustrophobia), VR solves a real logistical problem: the clinician does not have to take the patient to the airport or wait to find a suitable dog on the street. For social phobia, it offers graded audiences in privacy. For panic disorder with agoraphobia, it allows exposure to interoception and spatial contexts under controlled conditions. In all these cases, VRET is one more tool in the therapeutic repertoire, not a substitute for the cognitive, behavioral, and interpersonal work that underpins any solid intervention.

If you work with dog phobia, acrophobia, or claustrophobia, you can review our overview of the dog phobia exposure scenario, the panoramic glass elevator scenario for combined acrophobia and claustrophobia, and the clinical elevator scenario. To dig deeper into how presence is measured in these environments and why it influences treatment response, we recommend reading the article on presence and immersion in clinical VR.

If you want to explore VRET in your own practice, you can book a guided demo with the clinical team to review the available scenarios and discuss how they fit into your caseload.

This article is for informational purposes for psychology professionals. It is not clinical advice for any individual case and does not replace the judgment of the licensed psychologist in charge. VRET is professional clinical-support software, not a CE-marked medical device.