The Internet offers a new way of providing psychological treatment for common mental health problems [1]. Internet treatment has been described as a treatment that is operationalized and transformed for delivery via the Internet [2]. They are usually highly structured, and self-guided or partly self-guided with an identified therapist [3]. Internet interventions may however also resemble traditional psychotherapy with scheduled sessions [4], but overall a characteristic feature is that that patients are reached from a distance and that therapist time is reduced compared with face-to-face treatment [5]. Several independent research groups have developed Internet-based cognitive-behavioral interventions [6, 7], and randomized controlled trials suggest that Internet-based treatment with minimal therapist contact via e-mail can be effective in treating various conditions [8], including depression [9] with moderate to large effect sizes. Although several studies have shown that guided Internet treatment of depression [10] can lead to reduced symptoms, less is known about how participants experience these treatments. Reviews have however been conducted on the reasons why people fail to adhere to computerized interventions [11, 12]. A few qualitative studies have investigated experiences of guided self-help [13], including computerized treatments [14], showing that both advantages and disadvantages are reported by patients who have received guided self-help for depression. One study investigated experiences of an Internet depression treatment that was delivered in real time with scheduled online session appointments [15]. Results revealed two themes: developing a virtual relationship with a therapis and the process of communicating thoughts and emotions via an online medium. There is also a recent study on unguided Internet-delivered depression treatment [16] which concluded that the program tested did not work and that provision of support could improve outcomes.

Studies using qualitative methods can be helpful when investigating the potential benefits and disadvantages of treatment and explorative methodology of participants' experience can help us understand the mechanism of change. Since relatively little is known regarding predictors and mediators of change in Internet treatments [17], qualitative research may be justified help identify the unique aspects of Internet treatment. To our knowledge, no study has to date investigated how participants experience guided Internet-based self-help treatment for depression. This study therefore used qualitative methods to explore participants' experience of an Internet administrated guided self-help treatment for depression. We conducted the study alongside a randomized controlled trial [18], in which we investigated two forms of Internet treatment for major depression: e-mail therapy and guided self-help [19]. The overall aim of the study was to obtain a detailed understanding of participants' experience of the treatment that could contribute to improve the effectiveness and reduce dropouts in Internet-based depression treatments.

The aim of this study was to explore participants' experience of Internet administrated treatment for depression using qualitative methods. The analysis of participants' accounts of the treatment sixth months after the treatment yielded three distinct patterns of change. These related to participants' motivational experience of the treatment, how they worked with the treatment material, their attitudes towards the treatment, and their perception of what skills and knowledge they had gained from the treatment.

Overall, the results correspond with existing theoretical models of change in traditional face-to face psychotherapy for depression within the qualitative process research [28, 29]. The three change processes are partly in line with the three patterns of change described in study of clients' experience of a face-to-face short-term counselling [30]. The current study also indicates that therapeutic work in an Internet-based treatment is as a dynamic process, and that the treatment is perceived differently depending on expectations and outcome. This is in line with a previous qualitative study on guided self-help [13], and another qualitative study on how a mental health self-help clinic was perceived [31]

Although more research is needed to draw definitive conclusions, this study offers some clinical implications of value to those developing Internet-based self-help treatments. The results indicate that participants who reported that they only worked sporadically with the treatment, alternatively that they did not want to or could not work practically with the treatment material, were less inclined to view the treatment positively and also reported a less favourable outcome of the treatment. In comparison, participants who had a practical "hands-on" approach to work (testing and applying treatment strategies), expressed that they preferred the treatment format and had integrated the treatment principles in their everyday life. Redefining obstacles in treatment, confronting challenges, believing in their own ability, and acceptance of depression and of themselves also seemed to be significant themes within the narratives of participants who expressed improvement and a favourable outcome. Those themes seemed to be consistent with the findings in previous research on successful therapeutic work [28, 29].

Participants also gave different accounts in relation to the three change processes of how they perceived the support and the motivational aspects of the treatment. Some appreciated that they had to work on their own and felt that they had a "back up" support system, while others missed real-life conversation or felt pressured by the treatment program. This is partly in line with previous research which showed that an intense relationship with the therapist resulted in interrupted change patterns in time-limited counselling [30]. Our results indicate that both too much contact with the therapists (in particular if it is perceived as inadequate) and a lack of contact can both be hindering factors for treatment success. While participants may need some form of support in Internet administrated treatments [32], the current study also points to the importance of autonomy and feelings of self-efficacy in successful therapeutic work. In line with previous research [33], the findings highlight that an Internet program should take into account the non-specific factors of therapeutic work by developing a sufficient amount of support [9]. But in doing so, one should not jump to the conclusion that much more support will necessarily lead to better outcome. It might be that too much contact from the treatment program render loss in self-efficacy and make the participant believe that he or she is not the agent in the treatment or alternatively that the loss of face-to-face contact becomes more apparent when the contact is via electronic media [15]. Indeed, some expressed that they missed conversation and needed additional support, which might indicate that contact (e.g. via e-mail or face-to-face) is important to a varying degree depending on the participant. Individuals who have the ability and motivation to work on their own and have a structured and practical approach to work, are perhaps more suited for this form of treatment than others. More research is needed to determine if this is the case. On the other hand, a program designed to fit participants' motivation and ability (e.g. giving participants more time to complete treatment or more e-mail contact) might be suitable to increase adherence [34]. Indeed, tailoring of treatment could be feasible in particular in depression as comorbidity with anxiety disorders and somatic disorders is common [35].

This was an exploratory study and the interpretations of the data should be considered within the context of qualitative research [36]. The results can not be regarded as representative for all people who receive Internet administrated treatment for depression. The sample was limited, small and selected, and from a qualitative research point of view a larger number of participants would have been preferred given the way we analyzed our data as a fuller description might have emereged. Most importantly, the study included only one participant who had not improved at all from treatment (as measured with the CGI-I). If more such participants had been included the results might have been different. However, on account of this misrepresentation two more participants whose improvements were graded as minimally change were interviewed. All of the participants had an educational level of collage or higher, which may also limit how well our findings can be generalized. While qualitative research often involves relatively few participants, our sample may have been too small. Moreover, because the interviews took place six months after treatment had ended, the results rely on participants' memories of treatment, which could of course be subject to bias and also dependent on how well the treatment worked. This might make it difficult to draw any firm conclusions about sequences of change during treatment, participants' experiences of treatment content or attitudes towards treatment, because clients tend to construct memories in accordance with their present situation [37].

Another limitiation of the study relates to the potential bias since two of the involved researchers had worked with the treatment trial. We tried to handle the bias by involving a researcher with no previous experience of Internet research and also included a fifth co-author to handle the possible bias. Moreover, we constantly reflected on the research process and our own role when coding and interpreting the interviews [38], and were careful ton include not only participants who were very pleased with the intervention but also persons who had not improved.

The findings correspond with existing theoretical models of face-to-face psychotherapy within qualitative process research. Persons who take responsibility for the treatment and also attribute success to themselves appear to benefit more from Internet-based treatment. Further qualitative research is needed to investigate how participants experience Internet-based treatment for other problems than depression. This might not only cast light upon how Internet based self-help treatments work but may also give additional information about the active ingredients and the mechanism of change in face-to-face psychotherapy, in particular in relation to common and specific factors.