Can therapy in virtuo be better than in vivo?

Bouchard and colleagues have investigated the use of virtual reality in the treatment of psychological disorders for over a decade. Their latest research report, published in this issue, ^{Reference Bouchard, Dumoulin, Robillard, Guitard, Klinger and Forget3} presents the results of a randomised controlled trial (RCT) that found virtual reality to be superior to in vivo exposure on self-reported outcomes of social anxiety, both post-treatment and at 6-month follow-up. Therapists rated the virtual reality as more practical and less burdensome than using real-life exposure stimuli. These results are striking for two reasons. First, the superiority of virtual reality contradicts other recent trials in social anxiety that showed in virtuo exposure to have inferior ^{Reference Kampmann, Emmelkamp, Hartanto, Brinkman, Zijlstra and Morina4} or similar ^{Reference Anderson, Price, Edwards, Obasaju, Schmertz and Zimand5} effects compared with in vivo exposure. Second, the difference in outcomes was sustained after 6 months, even though in many comparative trials the gap between two effective therapy modalities – not only exposure – tends to narrow and even diminish at medium- to long-term follow-up.

The observed large effect of in virtuo exposure in Bouchard et al's study is attributed by the authors to two reasons. First, there was direct contact with a therapist during the in virtuo exposure tasks, in contrast to previous studies ^{Reference Kampmann, Emmelkamp, Hartanto, Brinkman, Zijlstra and Morina4,Reference Anderson, Price, Edwards, Obasaju, Schmertz and Zimand5} in which patients and therapists were in two separate rooms and patients engaged only in dialogues with virtual characters. Second, therapists used the virtual reality to delve into the patient's exposure experiences more that they were able to do in real life. Either way, a meaningful interaction with a therapist seems to be an important active ingredient for in virtuo exposure. Therapeutic guidance is thought to be a significant modifier for patient engagement and outcomes in other technology-mediated mental health interventions. ^{Reference Cuijpers, Marks, van Straten, Cavanagh, Gega and Andersson6,Reference Baumeister, Reichler, Munzinger and Lin7} Clinicians, researchers and patients often express mistrust and antipathy for technology when it is perceived as a threat to the ‘real’ therapist–patient interaction and when computer-generated responses of empathy – ‘am sorry to hear this’ – feel patronising and false. ^{Reference Gega, Smith and Reynolds8} Similarly, virtual reality's added value is not in imitating therapists, but in allowing therapists to maximise or expedite the returns from their interactions with patients.

Given that the aim of therapy is to improve people's lives and that virtual reality gains are generalisable, ^{Reference Opris, Pintea, García-Palacios, Botella, Szamosközi and David2,Reference Morina, Ijntema, Meyerbröker and Emmelkamp9} we would expect that patients are able, willing and encouraged to apply in real life what they learn in therapy. In Bouchard et al's study we infer that real-life application happened post-treatment, either spontaneously because patients improved, or as part of their relapse prevention plan. This may account for the sustained effects of in virtuo exposure at 6-month follow-up. Interestingly, to minimise the contamination of the ‘virtual’ with the ‘real’ during treatment, Bouchard et al discouraged real-life self-directed practice – in effect, exposure ‘homework’ – between in virtuo therapy sessions. This was similar to other virtual reality trials in social anxiety that either discouraged real-life exposure homework altogether ^{Reference Kampmann, Emmelkamp, Hartanto, Brinkman, Zijlstra and Morina4} or incorporated non-exposure homework (such as practising in front of the mirror or keeping a diary). ^{Reference Anderson, Price, Edwards, Obasaju, Schmertz and Zimand5}

Homework in the form of self-directed exposure tasks between clinic sessions is common practice and an established therapeutic ingredient proven to enhance therapy outcomes. ^{Reference Kazantzis, Whittington and Dattilio10} Its absence did not seem to compromise the effects of in virtuo therapy in Bouchard et al's study, although we do not know whether effects could have been greater should homework had been encouraged between virtual reality sessions. If virtual reality can achieve the same or even better results without self-directed real-life exposure practice, it may be a way of engaging patients who are unable or unwilling to carry out homework and may be considered ‘unsuitable’ for CBT or drop out because of it. Having said that, in virtuo therapy should not be disconnected from reality, but help the transition from clinic to real life. Optimal delivery that is true to common practice and maximises the chances of improvement should encourage the use of real-life self-directed exposure practice, both as homework between virtual reality sessions and as relapse prevention after treatment.

Finally, Bouchard et al considered that a limitation of their study was that the virtual reality exposure scenarios were not matched to the in vivo ones, as had been done in previous trials; ^{Reference Kampmann, Emmelkamp, Hartanto, Brinkman, Zijlstra and Morina4,Reference Anderson, Price, Edwards, Obasaju, Schmertz and Zimand5} yet, this may be a strength that accounts for the large effect of their in virtuo exposure therapy. Reasonably, experimental trials aim to match in virtuo and in vivo exposure in content and structure so that any differences in outcomes can be attributed to the differences between the virtual and the real when all other things are equal. Yet, the whole point of using virtual reality is to harness its full potential and go beyond what we can do in real life. For example, virtual reality can facilitate multicontext, tailored, mass practice of extreme exposure scenarios in a short period of time; ^{Reference Gega, White, Clarke, Turner and Fowler11} matching it to real life would mean that patients only experience the quantity and nature of scenarios that are possible in vivo and nothing more. With the proviso that patients follow-up on what they learn in the virtual world by applying it to real life, the optimal way of using in virtuo exposure is not to match it to in vivo, but to make full use of the virtual reality features that can have added therapeutic value over and above real life.

Future directions for in virtuo therapy

We need to move beyond circumscribed experimental comparisons and towards large-scale pragmatic RCTs where in virtuo therapy is delivered (a) as intended to be in routine practice and (b) in an optimal way that harnesses the full potential of virtual reality and uses treatment protocols that reflect modern theory, up-to-date evidence and crystallised clinical experience about what is likely to work best. First, both in vivo and in virtuo bona fide exposure therapies need to include response prevention (i.e. refraining from using safety behaviours to cope with anxiety) as in Bouchard et al, ^{Reference Bouchard, Dumoulin, Robillard, Guitard, Klinger and Forget3} and not encourage safety behaviours as in Freeman et al. ^{Reference Freeman, Bradley, Antley, Bourke, DeWeever and Evans12} Second, we must allow for a mismatch between in virtuo and in vivo exposure tasks in terms of content, intensity and quantity, because only then we can make full use of what virtual reality can offer that real life cannot. Third, in virtuo therapy needs to integrate real-life self-directed exposure practice as homework between sessions and as relapse prevention post-treatment, given that this is true to common practice and important for maximising and sustaining outcomes. Finally, in situ direct dialogues with a therapist delving into the patient's virtual experiences can optimise learning from the CBT adage that the way we respond to and think about an experience shapes our feelings towards it.

Meta-analyses have shown that virtual reality's effects are similar to conventional treatments and can be transferable and generalisable to real life. ^{Reference Valmaggia, Latif, Kempton and Rus-Calafell1,Reference Opris, Pintea, García-Palacios, Botella, Szamosközi and David2,Reference Morina, Ijntema, Meyerbröker and Emmelkamp9} Future non-inferiority pragmatic trials are best placed to confirm whether, within an acceptable margin of difference, in virtuo exposure is no better or worse than in vivo. Following on from this, the added value of virtual reality as a therapeutic tool hinges on its versatility, efficiency and ability to go beyond what we can do in real life, as well as on its potential to engage populations who for various reasons cannot benefit from conventional treatments. Virtual reality opens up a world of advanced therapeutic possibilities, but it is still far from influencing care pathways and improving patient outcomes in our day-to-day clinical practice. Despite the virtues of virtual reality, the proof of the virtual pudding is in the real-life eating.