The literature search identified 1637 papers once duplicates were removed, with a total of 69 papers reporting on 54 discrete studies meeting inclusion criteria (see Table 1). These included 21 randomized controlled trials (RCTs), 20 pre-to-post and non-randomized comparison trials, 6 case series, 4 stand-alone qualitative studies, and 3 studies examining rates of uptake. Of the RCTs, 7 examined efficacy compared with a non-therapy control, and 15 included a head-to-head comparison with in-person therapy, of which 9 conducted formal non-inferiority or equivalence analyses (detailed in Table 2). No studies contrasted videoconferencing with other remote communication modalities (e.g., telephone).

The most frequently studied diagnostic groups were post-traumatic stress disorder (PTSD; 14 studies), and depression (10), for which there were a number of well-powered RCTs, followed by anxiety disorders (6), obsessive–compulsive spectrum disorders (6) and eating disorders (6). Twelve additional studies examined mixed diagnosis populations including a large RCT. Across these studies, a number examined implementations to specific populations, with a large number, particularly PTSD studies, conducted with veterans or military personnel, and others focusing on populations with difficulties attending clinic settings in person, including people with difficulties leaving the home, people living in rural or remote areas, prison inmates, and geographically dispersed members of migrant populations. Less than half of studies were conducted within the person’s home/residence, with many especially older studies, involving visiting a local clinic using telehealth equipment to connect with a therapist in a different location. The types of technology used for videoconferencing included dedicated telemedicine hardware, analogue videophones, and, increasingly, using Internet-based videoconferencing software on computers or smartphones. Many studies provided participants with equipment such as laptop or tablet computers, but more recent studies have used participants’ own devices.

Across the full range of studies, therapy was found feasible to deliver via videoconferencing, clients were satisfied with therapy, and expected improvements in targeted symptoms occurred. We consider the findings for specific populations in detail (summarized in Table 3), followed by broader findings about use of videoconferencing across all studies.

Application of videoconferencing with different populations Post-traumatic stress disorder

PTSD was the most researched mental health diagnosis. In addition to small pre–post studies and pilot RCTs, the search identified 7 well-powered RCTs of videoconferencing therapy for PTSD, covering a range of treatment protocols, including cognitive processing therapy (CPT), prolonged exposure (PE), and behavioural activation.

Two trials examined the use of the eight-to-twelve session PE protocol to treat PTSD in veterans (Acierno et al., 2017, also reported on in Gros, Allan, Lancaster, Szafranski, & Acierno, 2018; Gros, Lancaster, López, & Acierno, 2018; and Yuen et al., 2015), and one trial combined behavioural activation with exposure therapy to treat both PTSD and depression (Acierno et al., 2016; Gros et al., 2012; Strachan, Gros, Ruggiero, Lejuez, & Acierno, 2012). All compared videoconferencing to in-person delivery and had samples that were over 90% male. Videoconferencing showed similar rates of therapy completion (Acierno et al., 2016, 2017; Yuen et al., 2015) and satisfaction (Gros, Allan, et al., 2018; Yuen et al., 2015) and was non-inferior to in-person for PTSD, depression, and anxiety (Acierno et al., 2016, 2017; Yuen et al., 2015).

Four trials examined CPT delivered by videoconferencing in comparison with in-person therapy (Glassman et al., 2019; Lui et al., 2019; Maieritsch et al., 2016; Morland et al., 2015). Participants were again predominantly veterans, with one study also including civilians (Morland et al., 2015), but females were better represented in CPT studies (Lui et al., 2019: 45% female, Morland et al., 2015: 100% female). Delivery by videoconferencing was found to be non-inferior to in-person in reducing PTSD symptoms in all studies other than Lui et al. (2019) who found that videoconferencing was inferior at post-treatment, but equivalent at 6-month follow-up. All studies found no significant differences in dropout or satisfaction between videoconferencing and in-person conditions.

Overall, the generally positive findings of acceptability and efficacy of videoconferencing for exposure-based therapies are noteworthy, suggesting this modality is able to support this emotionally challenging, experientially focused, treatment. It has also been observed that videoconferencing clients rate the therapeutic alliance as highly for exposure-based sessions as other CBT-based sessions (Germain, Marchand, Bouchard, Guay, & Drouin, 2010).

Depression

We identified 3 well-powered RCTs of videoconferencing therapy for depression, 3 smaller RCTs, and 4 studies using other designs. Two studies, including one of the RCTs (Yang, Vigod, & Hensel, 2019), primarily reported on uptake of videoconferencing. Intervention models included problem-solving therapy, behavioural activation and combined CBT protocols for depression with insomnia, and for depression with self-harm. Overall, results suggested participants were satisfied with therapy, and ratings of acceptability and efficacy appeared similar to in-person delivery.

Problem-solving therapy was examined in a three-arm RCT which compared videoconferencing or in-person delivery with a supportive weekly care-call control condition in 158 housebound adults over the age of 50 with depression (Choi, Hegel, et al., 2014; Choi, Marti, et al., 2014). On the Hamilton Rating Scale for Depression (HRSD), both videoconferencing and in-person problem-solving therapy were superior to the control condition at 12 weeks, 24 weeks, without differing from each other, and videoconferencing was superior to both conditions at 36 weeks (Choi, Marti, et al., 2014).

Videoconferencing-based behavioural activation has been examined in two RCTs, both conducted with veterans. Luxton et al. (2016) conducted an RCT of an 8-session behavioural activation intervention delivered by telehealth or in-person to 121 military personnel and veterans with depression. Both conditions showed significant post-treatment improvements on the Beck Depression Inventory (BDI) as the primary outcome, and non-inferiority analyses showed videoconferencing was non-inferior at mid-treatment and 12-week follow-up, but not immediately post-therapy. Egede et al. (2015) obtained more conclusive results in a larger non-inferiority trial with 241 older veterans with major depression. Comparing videoconferencing delivery using a videophone system with in-person delivery, non-inferiority was established with no significant differences observed in trajectories of improvement on the BDI and Geriatric Depression Scale, with rates of recovery similar between conditions.

Smaller studies have additionally demonstrated feasibility and acceptability of delivering CBT-based therapies via videoconferencing to specific populations such as women with post-partum depression or anxiety (Yang et al., 2019) and Korean migrants with depression (Jang et al., 2014). Among other notable studies, Scogin et al. (2018) conducted a small RCT of a 10-session CBT-based treatment for comorbid depression and insomnia delivered via Skype, which found superiority over usual care on a measure of insomnia, but not the HRSD. Finally, in treating self-harm, Sayal et al. (2019) commenced a small RCT (N = 22) of problem-solving therapy for young adults following presentation for self-harm and mild depression. However, this was discontinued due to recruitment difficulties (an analysis of which did not attribute these to the use of videoconferencing).

Anxiety disorders

Anxiety disorders have been less fully studied than depression. Nonetheless, anxiety disorders feature as a major group in a number of mixed diagnosis studies, which have demonstrated that CBT-based therapies can be satisfactorily delivered (e.g., Brunnbauer et al., 2016; Dunstan & Tooth, 2012; Griffiths, Blignault, & Yellowlees, 2006; Matsumoto et al., 2018, 2020; Stubbings et al., 2013). Among these, an RCT design was used by Stubbings, Rees, Roberts, and Kane (2013) in a study of 26 people with mainly anxiety disorders. Reductions on all subscales of the Depression Anxiety Stress Scale (DASS) were observed following videoconferencing CBT, and, while underpowered, no differences in the magnitude of effect were observed between videoconferencing and an in-person comparison group. The feasibility of applying videoconferencing to deliver therapies to other specific populations is indicated by the following, mainly small, studies.

Generalized anxiety disorder (GAD)

A multiple baseline case series by Théberge-Lapointe, Marchand, Langlois, Gosselin, and Watts (2015) showed evidence for successful cognitive behavioural treatment of GAD, with five participants no longer meeting diagnostic criteria post-therapy and 3 months later, and this outcome persisting to 12 months after treatment in all but one case. At the time of writing, initial results from a large RCT of CBT for GAD (N = 115), focusing on working alliance, have been reported by Watts et al. (2020), with clients rating the working alliance more highly for videoconferencing than in-person therapy across time points, although therapists rated both modes of delivery similarly.

Panic disorder and agoraphobia have only been studied in small pre-to-post studies, all of CBT. Bouchard et al. (2000) found significant improvements across all measures, reporting that five out of the eight participants no longer experienced panic attacks after the 12-week treatment. Bouchard et al. (2004) delivered the same intervention to a further 10 videoconferencing cases, compared with a non-randomized in-person delivery group. Nearly all participants achieved remission at the end of treatment, maintained six months later, a similar to in-person delivery. Matsumoto et al. (2018) also found significant reductions in panic symptoms among 10 participants with panic disorder in their pre-to-post study of CBT.

Social anxiety

Modality of delivery is of particular interest for social anxiety, where communication itself is a source of anxiety. Yuen et al. (2013) examined 12 sessions of acceptance-based behaviour therapy for 24 individuals with SAD. Therapists rated the use of videoconferencing as feasible, and there were post-therapy improvements on several questionnaire measures of social anxiety, maintained and at the 3-month follow-up, as well as changes on observer-rated social behaviour; participants indicated that they were satisfied with the treatment. Likewise, Matsumoto et al. (2018) found reductions in social anxiety following videoconferencing-based CBT in their small sample of 10 social anxiety participants.

Health anxiety

The largest study for a specific anxiety disorder has been for health anxiety: Morris et al. (2019) conducted an RCT comparing CBT delivered via videoconferencing or telephone with routine care in 156 participants. Supporting the use of videoconferencing, health anxiety was reduced in the therapy group relative to routine care at 6-, 9-, and 12-month time points.

Obsessive–compulsive and related disorders

Research into videoconferencing-delivered psychological treatments in obsessive–compulsive and related disorders was limited, with studies limited to case series and small sample single-arm open trials and pilot RCTs. Nevertheless, there is an emerging support for the acceptability and effectiveness of videoconferencing for a range of intervention types across OCD, hoarding and trichotillomania.

Matsumoto et al. (2018) reported on a standard 16-week CBT treatment for their 10 OCD patients. Symptom reduction pre–post treatment, strong therapeutic alliance, high rates of satisfaction with treatment, and 100% retention, supported the effectiveness and feasibility of the intervention. Further, two studies (Goetter, Herbert, Forman, Yuen, & Thomas, 2014; Vogel et al., 2014) successfully used exposure and response prevention (ERP) to treat OCD via videoconferencing, with post-treatment symptom reductions. Vogel et al. (2014) noted high engagement with treatment, an ability to observe exposure exercises as they occur in participants’ natural environments, and an opportunity to involve family members and carers, thus addressing family accommodation to rituals where appropriate.

Emerging investigations in hoarding and trichotillomania provide support for its effectiveness, feasibility, and that it provides additional benefits when compared to existing treatments. Muroff and Steketee (2018) delivered a structured CBT treatment for seven patients with hoarding. Six of the seven patients experienced improvements in symptoms post-treatment, with five maintaining the gains at 3-month follow-up. The ability to use portable devices to move around rooms was noted as an important facilitator in the treatment. In relation to trichotillomania, Lee, Haeger, Levin, Ong, and Twohig (2018) conducted an RCT comparing videoconferencing-based ACT-enhanced Habit Reversal Therapy to waitlist control in 22 trichotillomania patients. The study had high retention rates with only one dropout in each condition, and high levels of participant satisfaction and therapeutic alliance. Statistically and clinically significant improvements in trichotillomania symptoms were noted among the treatment group.

Eating disorders

In the treatment of eating disorders, there has been a single large RCT, which examined CBT for bulimia nervosa and related disorders (Ertelt et al., 2011; Marrone, Mitchell, Crosby, Wonderlich, & Jollie-Trottier, 2009; Mitchell et al., 2008). Although bulimia symptoms reduced for both videoconferencing and in-person delivery, and rates of abstinence from bingeing and/or purging showed were similar, the reduction in binge eating frequency was less for videoconferencing participants across multiple time points (Mitchell et al., 2008). Working alliance was rated similarly by clients for each of the conditions, but therapists rated the alliance less strongly in the videoconferencing condition (Ertelt et al., 2011).

Most other studies identified by the search examined smaller single group samples for bulimia and related disorders, reporting reductions in bulimic symptoms (Abrahamssom, Ahlund, Ahrin, & Alfonsson, 2018; Hamatani et al., 2019; Simpson et al., 2006) and satisfaction with the online modality (Abrahamssom et al., 2018; Simpson et al., 2005).

For anorexia nervosa, Giel et al. (2015) conducted a single group pilot study examining a relapse prevention intervention based on the Maudsley model (Schmidt, Magill, & Renwick, 2015) in 16 individuals. Eight sessions were delivered via videoconferencing, bookended by two in-person sessions. Three-quarters of participants completed therapy, rating high satisfaction, and at post-intervention body mass index had increased by an average of 1.1 points, eating concerns were reduced, and two participants were in complete remission.

Other populations

No studies were identified providing data on videoconferencing therapy delivery to persons with psychotic disorders, bipolar disorder, or personality disorder.

Client and practitioner experience Overall acceptability

Every RCT comparing at-home videoconferencing with in-person delivery at a clinic reported no group differences on questionnaire measures of satisfaction (see Table 2). Differences in satisfaction or dropout were only seen in two studies overall, both delivering interventions within the same environment: Morland et al. (2015) reported lower satisfaction ratings primarily related to negative experiences of the clinic setting that was attended for videoconferencing (also used in the in-person condition), suggesting specificity to the potentially impersonal experience of attending a clinic for a video-based appointment. Conversely, Choi, Hegel, et al. (2014) found that housebound people with depression receiving in-home therapy via videoconferencing were more satisfied than those being visited by a therapist. Overall, this demonstrates that satisfaction with videoconferencing-based therapy is as high as traditional forms of delivery.

In terms of therapy dropout, nearly all comparisons with in-person therapy revealed no group differences (see Table 2). An exception was a follow-up analysis of discontinuation in the trial by Acierno et al. (2017) reported that early dropout tended to arise more often with videoconferencing (Gros, Allan, et al., 2018), even though overall session attendance rates were similar. While dropout seems to only arise in a small number of people, other studies report discomfort with videoconferencing being cited by participants as a reason for dropout, so this may be an issue with a small number of people, although at this stage there is a lack of information on what contributes to this (Germain, Marchand, Bouchard, Drouin, & Guay, 2009; Lichstein et al., 2013; Simpson, Bell, Knox, Mitchell, & Eating, 2005).

It should be noted that individual comments expressing a preference for in-person therapy were often noted from videoconferencing participants (Choi, Wilson, Sirrianni, Marinucci, & Hegel, 2013; Lichstein et al., 2013; Simpson et al. 2005, 2006). Among qualitative client reports, a period of early discomfort and adaptation to using videoconferencing technology was also an experience reported by participants across studies (Choi, Hegel, et al., 2014; Dunstan & Tooth, 2012; Fitt & Rees, 2012; Germain et al., 2009; Lichstein et al., 2013; Simpson et al., 2005, 2006; Simpson et al., 2015; Yuen et al., 2015). For some participants, attitudes towards videoconferencing (including scepticism, anxiety, unfamiliarity) were linked to the experiences of discomfort in early sessions (Arnaet, Klooster, & Chow, 2007; Choi, Hegel, et al., 2014; Fitt & Rees, 2012; Simpson et al., 2005, 2006). For most, this early discomfort was reduced over time, as participants got more comfortable with the technology (Choi, Hegel, et al., 2014; Dunstan & Tooth, 2012; Simpson et al., 2005, 2006; Simpson et al., 2015) or their interactions with their therapist became more ‘natural’ (Yuen et al., 2015), although this did not always occur (Choi, Hegel, et al., 2014; Choi et al., 2013; Lichstein et al., 2013; Simpson et al., 2005). Therapists reported similar experiences of initial apprehension and discomfort, before becoming more confident in using videoconferencing technology and adapting to the modality (Dunstan & Tooth, 2012; Michell et al., 2008). This is balanced by other reports of participants embracing the novelty and use of technology in therapy delivery (e.g., Aranaet et al., 2007; Choi, Hegel, et al., 2014; Choi et al., 2013; Dunstan & Tooth, 2012). It should be noted that many of these studies were conducted before the widespread day-to-day use of videoconferencing platforms, and less adaptation may be required in the 2020s.

Facilitating access

One of the presumed benefits of videoconferencing is that it facilitates access. As shown in Table 1, many of the studies reviewed targeted participants in rural or geographically remote areas, and some involved applications to potentially isolated groups (e.g., housebound older adults; victims of domestic violence, migrants). Participant reports indicated that many people receiving videoconferencing therapy would otherwise have been unable to access any therapy (Choi et al., 2013; Hassija & Gray, 2011), while others included references to challenges of travel distance and its associated financial impact (Abrahamsson et al., 2018; Simpson et al., 2005; Simpson et al. 2015). Some studies also referred to the opportunity to provide specialist services for a specific issue to people over a broad area (Hassija & Gray, 2011; Lee et al., 2018).

Even when not an absolute barrier, the increased accessibility appeared valued. The post-partum mental health study by Yang et al. (2019) examined uptake when the option to use videoconferencing in place of in-person psychotherapy sessions was offered: 74% used videoconferencing for at least one therapy session, with 21% doing all therapy via video; Time and cost savings were identified, and participants reported being able to attend more frequently. In other studies, participants spoke of convenience, such as fitting therapy into busy life schedules (Abrahamsson et al., 2018; Choi, Hegel, et al., 2014; Choi et al., 2013; Lee et al., 2018; Yuen et al., 2015), and being able to access therapy from home (Choi, Hegel, et al., 2014). Continuity of care independent of location was also highlighted, both in relation to moving house (Simpson et al., 2005, 2006), and being released from prison (Morgan, Patrick, & Magaletta, 2008).

Symptoms of anxiety, concerns about stigma, and negative thought processes also featured as potential barriers to accessing in-person services that videoconferencing was able to circumvent (Abrahamsson et al., 2018; Bouchard et al., 2000; Simpson, Guerrini, & Rochford, 2015). For example, in the trichotillomania study by Lee et al. (2018), 40% of participants reported that they would not have entered treatment in an in-person setting due to shame. Privacy for persons in small or rural communities was also referred to (Simpson et al., 2005; Simpson et al., 2015). Nonetheless, privacy was not always assured by videoconferencing with concerns about privacy from others within the person’s own home being raised by some participants (Abrahamsson et al., 2018; Choi, Hegel, et al., 2014; Franklin, Cuccurullo, Walton, Arseneau, & Petersen, 2017). Notably, concerns about privacy from use of networked digital technology did not tend to be reported.

Client factors predicting uptake and satisfaction

Studies of client variables predicting uptake, engagement and completion of therapy have identified relatively few predictors. In considering predictors of uptake among American primary care attendees with a positive depression screen, Deen, Fortney, and Schroeder (2013) found that uptake of videoconferencing-based CBT was predicted by perceiving illness to be persisting, believing that treatment would be effective, and reporting geographic barriers to attending; Time barriers, financial barriers, perceived stigma, and other beliefs about depression were unrelated to uptake. In a mixed diagnosis veteran sample offered therapy, Valentine et al. (2020) found that videoconferencing therapy uptake, and sessions completed, were each unrelated to age, race, gender, and marital status.

Several studies have examined predictors of differential satisfaction with, or dropout from, videoconferencing therapy. In most studies, completion of therapy appears unrelated to baseline demographic (age, gender, ethnicity, income) and clinical variables (Choi, Hegel, et al., 2014; Germain et al. 2009; Luxton et al., 2016; Watts et al., 2020), although unreplicated findings reported by single studies include greater completion rates for mood rather than anxiety disorders (Valentine et al., 2018), lower baseline PTSD and absence of disability status (Gros, Allan, et al., 2018), and, among veteran samples, being an older, Vietnam-era veteran (Gros, Yoder, Tuerk, Lozano, & Acierno, 2011). Pruitt et al. (2019) also found that satisfaction with therapy was higher for older military, although, in their sample, this was confounded with serving vs veteran status, with active military needing to travel off base to access videoconferencing facilities. Analyses of predictors of outcome have been limited, but in a military sample, Smolenski, Pruitt, Vuletic, Luxton, and Gahm (2017) found greater baseline anxiety and loneliness predicted participants having a better outcome from in-person than from videoconferencing-based therapy.

Among their participants with depression, Choi, Hegel, et al. (2014) found no relationship between ratings of treatment acceptability and computer/Internet ownership, or network quality. Similarly, in an analysis of PTSD trial data, Price and Gros (2014) observed that outcome of PTSD treatment via telehealth was unrelated to prior experience with, or expressed comfort with, telehealth at the outset of treatment. This suggests that prior experience is not a requirement to benefit. However, prior experience of therapy appears to predict completion. In their study of uptake, Deen et al. (2013) found predictors of treatment completion were different from those for uptake, and completion was most related to engagement with other treatments: receipt of prior counselling and being prescribed antidepressant medication. Watts et al. (2020) also found prior therapy experience predicted completion.

Technical issues

Most studies referred to technical issues as an experience impacting on the delivery of therapy. These included difficulties establishing connection, disconnection, suboptimal audio and visual quality, and bandwidth and connection stability issues resulting in lag and frozen images. Participants considered minor disruptions such as lag as a frustrating and distracting disadvantage of videoconferencing, but, overall, this did not negatively impact on participant engagement (Abrahamsson et al., 2018; Choi et al., 2013; Dunstan & Tooth, 2012; Lichstein et al., 2013). Studies resolved these issues through in-session troubleshooting or reconnection. Severe technical issues (e.g., disconnection and inability to re-establish connection) were managed by postponing or cancelling scheduled sessions, or by utilizing a back-up communication method (e.g., telephone) (Abrahamsson et al., 2018; Germain et al., 2010; Hassija and Gray, 2011; Lee et al., 2018; Luxton, Pruitt, O'Brien, & Kramer, 2015; Olden et al., 2017; Yu et al., 2020; Vogel et al., 2014; Watts et al., 2020).

To proactively manage technical issues, test calls or in-person training were often provided to therapists to resolve potential technical issues at the outset (Acierno et al., 2016, 2017; Choi, Hegel, et al., 2014; Choi, Marti, et al., 2014; Goetter et al., 2014; Gros, Allan, et al., 2018; Gros, Lancaster, et al., 2018; Luxton et al., 2015, 2016; Yuen et al., 2013, 2015), and many studies arranged for technical support to be available as part of the study design (Acierno et al., 2016, 2017; Germain et al., 2009, 2010; Liu et al., 2019; Olden et al., 2017; Scogin et al., 2018; Watts et al., 2020; Yuen et al., 2013). Yuen et al. (2013, 2015) observed that technical difficulties reduced over the course of the study, in part due to participants becoming proficient at troubleshooting. Overall, while technical issues were encountered in most studies, participant feedback and reports from the study authors indicate that disruptions were not sufficiently impactful to detract from therapy.

Therapy relationship and process

On both formal measures and in qualitative reports, studies consistently reported that the videoconferencing clients were typically able to develop a positive connection with the therapist (Simpson et al., 2005, 2006; Simpson et al., 2015; Choi et al., 2013; Dunstan & Tooth, 2012; Fitt & Rees, 2012; Yuen et al., 2015), although some individual reports found a reduced sense of the therapist’s presence (e.g., Arnaet et al., 2007; Choi, Marti, et al., 2014). Furthermore, nearly all well-powered RCTs that directly compared client ratings of the therapeutic relationship with in-person delivery found no significant differences (Table 2), consistent with observations in smaller studies (e.g., Morgan et al., 2008; Scogin et al., 2018). Additionally, in an analysis of equivalence, Maieritsch et al. (2016) found confidence intervals for the working alliance fell within a priori bounds of equivalence in their trial of CPT. An exception to these findings is the CPT trial by Morland et al., (2015), which found statistically, but marginally, lower ratings for videoconferencing in the second session, with no differences at later time points. It is notable that, mirroring the adaptation to discomfort reported in some studies, some studies have also observed that videoconferencing clients rate a stronger alliance as sessions progress (Ertelt et al., 2011; Germain et al., 2010).

The converse finding of a stronger alliance in the videoconferencing condition by Watts et al. (2020) corresponds to qualitative comments in other studies to there being potential advantages of videoconferencing for the therapeutic relationship. Participants discussed finding therapy easier through having a greater sense of control (i.e., of emotion, of context, of the ability to leave) and the creation of a less intense therapy environment (Dunstan & Tooth, 2012; Fitt & Rees, 2012; Simpson, 2001; Simpson, Deans, & Brebner, 2001, Simpson et al. 2005; Simpson et al. 2006; Simpson et al. 2015). Participants discussed the ability to ‘talk more freely’, being less self-conscious, finding it easier to communicate and feeling less pressured or intimidated in videoconferencing than they might be in-person (Fitt & Rees, 2012; Simpson et al., 2005, 2015; Yuen et al., 2015).

Two of the three well-powered studies that included both client and therapist ratings (Ertelt et al., 2011; Morland et al., 2015; Watts et al., 2020) identified differences of perspective: videoconferencing clients rated a stronger alliance than in-person clients, while therapists rated the conditions the same (Watts et al., 2020), or clients rated the conditions similarly when therapists rated videoconferencing as inferior (Ertelt et al., 2011). These quantitative findings correspond to therapist reports of some difficulties in detecting emotion and ability to read body language through videoconferencing (Dunstan & Tooth, 2012; Simpson et al., 2005; Yu et al., 2020; Yuen et al., 2013). This highlights that therapists and clients may have discrepant experiences of videoconferencing therapy and that therapists can find the process of therapy more challenging, without that necessarily being reflected in client experience.

Adaptations of therapy

To deliver therapy via videoconferencing, several studies reported adaptations to therapy protocols. Most commonly, the practical logistic changes to how components of therapy were delivered involved using other technologies to share documents (e.g., mailing, faxing, emailing, or screen sharing worksheets and homework) (Gros et al., 2011; Himle et al.,