Introduction: CBTpsych.com is a fully functional intervention that aims at bridging the gap between Eliza (Weizenbaum, 1966) and modern internet treatments for anxiety disorders (Helgadottir, Menzies, Onslow, Packman & O‟Brien, 2009a). A Phase I trial demonstrated that two participants no longer met the diagnosis of social phobia on the DSM-IV and ICD-10 after being treated by CBTpsych.com. The quality of the interaction appeared to be similar to face-to-face therapy. The automated techniques were successful in engaging the participants and in encouraging them to log on regularly and complete the treatment (Helgadottir, Menzies, Onslow, Packman & O‟Brien, 2009b). The current study is a phase II trial targeting social anxiety in stuttering.
Method: File audit data were gathered from 10 years of experience in the psychological management of social anxiety and stuttering (St Clare et al., 2008). This data was used to create a fully automated human-like intervention using algorithms alone. CBTpsych.com creates cognitive restructuring exercises, behavioural experiments, and negative thoughts checklists for the automated treatment program. Furthermore, other features such as imagery rescripting were incorporated to adhere to the Clark D. M. and Wells (1995) model of social anxiety treatment. 18 participants presenting for treatment of stuttering with social anxiety at the Australian Stuttering Research Centre (ASRC) were offered 5 months access to the “computer psychologist”. 16 participants started using the computer program. The treatment did not involve any contact with clinical psychologists at the ASRC.
Results: Post-treatment analysis revealed that of the 16 participants who started using the program, 78% no longer met the criteria for social phobia. Furthermore, none of the 10 participants who completed all 7 sections of the computer program within 5 months had social phobia diagnosis at post-treatment. Consequently, in this study the “computer psychologist” was able to identify the specific problem areas participants reported and to design individualized formulations and tailored treatment components with corrective feedback. Furthermore, participants were able to engage in highly specific cognitive restructuring exercises, and with the help of the computer psychologist managed to build behavioural experiments to test out unhelpful cognitions. Data from this Phase II trial suggest that the computerised treatment protocol can lead to substantial reductions in anxiety and avoidance, and improve mood and quality of life as measured by the FNE, SPAI, BDI-II, UTBAS, SASS and OASES. Conclusions: The computer psychologist motivated people both to log on regularly (with use of an automated email process) and to complete the program. It engaged clients in detailed cognitive restructuring work, creating successful behavioural experiments for exposure, and producing promising preliminary results. However, these results need to be tested in a randomized controlled trial to establish efficacy. A RCT is currently under way. The preliminary findings presented here suggest that it would be viable to develop this type of treatment intervention for other anxiety disorders. Limitations of this approach and suggestions for future research are discussed.