A Comprehensive Review and a Meta-Analysis of the Effectiveness of Internet-Based Psychotherapeutic Interventions
Azy Barak
Liat Hen
Meyran Boniel-Nissim
Naâama Shapira
The Internet has been used for psychotherapeutic interventions for more than a decade. Various terms have been used to denote this special professional activity: etherapy (or counseling), online therapy, Internet therapy, and cybertherapy, and sometimes it is referred to as e-health or telehealth, as a part of more general activities. Although attempts have been made to associate specific terms with more focused activities (e.g., cybertherapy for the use of virtual reality software), this terminology has failed in practice, and professionals and laypersons alike normally use different terms interchangeably. There are, however, several major factors that differentiate among the different therapeutic applications conducted by means of the Internet. One of these has to do with the online-intervention method employedâwhether it includes human communication (termed here etherapy) or is a self-help, website-based therapy (termed here web-based therapy). A different major factor has to do with another Internet-enabled capacityâwhether an intervention is delivered in âreal-timeâ (synchronously) or is delayed (asynchronously). A third important factor has to do with mode of communicationâwhether conducted textually, by audio only, or by video (webcam). Other important differentiations have to do with individual versus group mode and therapeutic approach, terms normally associated with traditional, face-to-face therapies.
From its start, Internet therapy has been criticized and opposed by both many laypeople (e.g., Skinner & Latchford, 2006) and professionals (e.g., Lester, 2006; Wells, Mitchell, Finkelhor, & Becker-Blease, 2007) on several grounds. First, the lack of face-to-face visibilityâwhich prevents the transmission and detection of a clientâs nonverbal communication cues, on the one hand, and the use of a therapistâs body language, on the otherâcreated massive resistance. This opposition based on the lack of nonverbal communication was considered an essential component of therapeutic relationships. Second, ethical issuesârelating to secrecy and confidentiality, identity of patients and therapists, impersonations, handling of emergency situations, and moreâbecame a central problem with the application of computer-mediated, distance therapy. Third, contemporary laws and regulations did not always cover various situations created by online therapy, such as local licensing requirements, legal jurisdiction, professional insurance of negligence, and more, resulting in often unresolved legal issues. Fourth, practical and technical concerns led to arguments related to the training of online therapists, the dependency on electricity and on complicated, fragile technologies, as well as to worries about the digital divide, and more. All of these criticisms, though they still exist, have been answered to a great degree as the field developed, numerous professionals have connected, literally and figuratively, to this new channel of therapy, advanced technologies emerged, ethical codes were developed, training courses and workshops began to be offered, and so on (Chester & Glass, 2006; Grohol, 2004). Not least, many clients seemed to like this innovative therapeutic option (King et al., 2006).
In the attempt to respond to the questions and critiques posed by the opponents of online therapy, quite a few therapeutic process studies were conducted that generally tried to examine the special characteristics of the therapeutic dynamics created in distant, invisible, interpersonal circumstances. The findings of those investigationsâobviously concentrating in etherapy of different formsâoften showed that therapy processes online are, in many ways, similar to the traditional form of therapy, though they possess some unique features as well, which were identified. Cook and Doyle (2002), for example, found that clients of e-mail- or chat-based therapy rated therapeutic working alliance similar and even superior to that of face-to-face therapy. In an analogue study, Mallen, Day, and Green (2003), however, found higher ratings of disclosure, closeness, and satisfaction with the face-to-face therapy experience than with that online, though no difference in emotional understanding was detected between the two interaction modes. Escoffery, McCormick, and Bateman (2004) reported on the process development of goal-setting, consciousness growing, and satisfaction of clients using web-based therapy for smoking cessation. Lewis, Coursol, and Herting (2004) studied client and counselor experiences in videoconferencing-based counseling in a qualitative analysis of a case study. They were able to qualify interesting themes in counselorâs and clientâs experiences, and noted the clientâs positive feelings. In an analogue study, Rochlen, Land, and Wong (2004) found that men with high emotionality preferred online counseling over face-to-face counseling more than did men with low or restricted emotionality. Bickmore, Gruber, and Picard (2005) showed that bond and working alliance may be achieved even when working with an automated software agent. In an observational analogue study, Rees and Stone (2005) saw that clinicians rated working alliance in videoconferencing-based therapy lower than in traditional, face-to-face sessions. Young (2005) investigated the attitudes of clients treated through online chat groups; although convenience and anonymity were cited as favorable factors, privacy and security concerns were listed against its use. Barak and Bloch (2006) found that session-impact factors in chat-based therapy were related to the perceived helpfulness of sessions. Leibert, Archer, Munson, and York (2006) studied clientsâ working alliance in, and satisfaction with, e-mail- and chat-based counseling; both were rated inferior to face-to-face experiences. Reynolds, Stiles, and Grohol (2006) found session-impact factors and therapeutic alliance in e-mail-based therapy to be similar to face-to-face therapy for both therapists and clients. According to Ritterband et al. (2006), the use of audio, graphics, and interactivity in website-based treatment of encopresis with children contributed to their elevated knowledge, motivation, and readiness to change. Thus, generally speaking, these studies show that counseling and psychotherapy relationships can effectively take place under the special circumstances enabled by the Internet as far as major therapeutic processes are concerned.
The main questions consistently asked throughout these studies and through numerous other publications have been whether therapy practiced online was effective, whether therapy could be conducted effectively (i.e., achieve its therapeutic goals) through the Internet, whether it was as effective as traditional therapy, and how various methods and variables associated with online therapy affected its effectiveness. Although quite a few individual studies on treating a variety of psychological problems have been conducted to date and numerous case studies have been published and presented, a comprehensive review and examination of these questions are still lacking.
Several attempts were made to provide an inclusive review on the effectiveness of online psychological interventions. These attempts, however, offered a rather limited view of the question at hand, mainly because of their partial inclusion of the published research (e.g., Anthony, 2006; Ritterband et al., 2003; Tate & Zabinski, 2004; Ybarra, Eaton, & Bickman, 2005); stenographic, encyclopedic-style summary (Barak, 2004); emphasis on history and development rather than effectiveness (e.g., Skinner & Zack, 2004); concentration on a specific problem area, such as anxiety (Andersson, Bergström, Carlbring, & Lindefors, 2005), depression (Andersson, 2006), panic disorder (Carlbring & Andersson, 2006), smoking cessation (Etter, 2006; Walters, Wright, & Shegog, 2006), weight loss (Weinstein, 2006), health-related problems (Strecher, 2007), or problem drinking (Walters, Miller, & Chiauzzi, 2005); concentration on web-based interventions only (Andersson, 2006; Griffiths & Christensen, 2006; Pull, 2006); focus on video-based therapy (Simpson, 2003); or their mixing together therapy and support (Mallen, Vogel, Rochlen, & Day, 2005). Although the general conclusion of these reviews, as well as several others, was highly supportive of Internet therapy, it seems that reliance on these resources is insufficient because their surveys of the literature are in effect incomplete or narrow. In addition, none of these reviews made an attempt to examine interactions of relevant moderators (e.g., age of clients, therapeutic approach) with therapy outcome. It should be noted that several books focusing on etherapy and e-counseling (e.g., Derrig-Palumbo & Zeine, 2005; Hsiung, 2002; Kraus, Zack, & Stricker, 2004; Tyler & Sabella, 2003) also provided a general and partial review of research, as well as numerous case examples, but did not provide a thorough and comprehensive view of the area.
Three meta-analytic reviews that are relevant in part to our current research questions were conducted. Wantland, Portillo, Holzemer, Slaughter, and McGhee (2004) conducted a meta-analysis of 22 web-based versus non-web-based psychological interventions intended to educate and create behavioral change in people with chronic illness. They found a large variability in effect size (ES), ranging from â0.01 to +0.75, which averaged out to a moderate mean ES. In another meta-analytic review, Spek et al. (2007) examined 12 studies that tested the effectiveness of web-based cognitive-behavioral therapy (CBT) for depression and anxiety. They found a small-to-moderate ES for the treatment of depression and a large ES for the treatment of anxiety. Provision of therapist support (provided online) moderated these findings, as therapist support resulted in large effects and no such support resulted in small effects. These two meta-analyses referred only to web-based interventions in specific problem areas. Hirai and Clum (2006b) conducted a meta-analysis of the effectiveness of various self-help venues in helping people with anxiety problems, including computer and Internet self-help interventions among other methods (e.g., printed materials, videotapes). They found that computer- and Internet-based self-help interventions appeared, for the most part, to yield equally effec...