In order to uncover the origin of Internet-based cognitive behaviour therapy (ICBT), it is important to acknowledge the different strands in the literature. One source of differentiation is choice of treatment format. There are two treatment formats of particular importance. The first is cognitive behaviour therapy (CBT), which is the most well researched form of psychotherapy. CBT is also increasingly used in health care, particularly for the treatment of mild to moderate psychiatric conditions (Andersson et al. 2005a). Arguably the most evidence-based psychological treatment, CBT is now disseminated on a wide scale in countries like the United Kingdom (Layard and Clark 2014). It is beyond the scope of this book to discuss CBT in detail, but we can outline a few of its key characteristics. The first step in CBT is a case conceptualisation, which can be more or less behavioural in its orientation. The next step is typically to present a treatment rationale to the patient. The subsequent treatment steps may vary depending on the agreed-upon treatment goals, but behaviour problems (too much or too little) and irrational beliefs are common targets, so specific treatment techniques have been developed to address them (Westbrook et al. 2011). These specific treatment techniques include behavioural activation in the treatment of depression and exposure when treating anxiety and avoidance. These techniques are framed using a rationale based on a cognitive and/or a behavioural conceptualisation. CBT is often focused on problems “here and now”, and recipients of CBT are instructed to practise techniques both during sessions and between sessions in the form of homework. The duration of the treatments may vary, but time constraints typically limit treatment to 10–20 weeks. However, there are examples of shorter treatments (like one-session treatments of specific phobia) and longer treatments (like dialectical behaviour therapy for borderline personality disorder, in which the first phase alone can last a year). Finally, relapse prevention and follow-up sessions may be included as well by the end of the treatment. While that concludes our overview of the execution of CBT, there are many more CBT techniques and procedures that we did not discuss, such as relaxation techniques, mindfulness, response prevention and specific cognitive techniques like Socratic questioning (this is sometimes, but not always regarded as a CBT technique). The key to CBT’s success is likely the fact that treatments are adjusted according to the problem at hand (Barlow 2004). Thus, CBT for chronic pain will differ from CBT for something else, such as post-traumatic stress disorder (PTSD), even if there is some overlap. Another factor that uniquely characterises CBT is the way it is delivered. Whereas standard psychotherapy is typically done in the form of individual face-to-face treatment, CBT has been conducted in many different forms, such as group, family and couple therapy. CBT has even been done in a large lecture format. However, most noteworthy to us is the vast literature on CBT as self-help (Watkins 2008).

CBT has been the topic of self-help research for a long time, which is often referred to as bibliotherapy (not to be confused with self-help groups led by patients themselves that are unrelated to CBT). Particularly, guided self-help using text material (mostly books) has been studied in controlled trials, with results showing moderate to large effects for a range of psychiatric and somatic conditions (Watkins and Clum 2008). Although ICBT was not developed until later, ICBT programmes are influenced by or even identical to the information presented in these evidence-based self-help books. There is a confusion in the literature because early bibliotherapy studies are sometimes collapsed with ICBT studies in reviews (Cuijpers et al. 2010). It is possible that guided bibliotherapy and guided ICBT are about as equally effective when therapist guidance is involved. For example, while two of the authors of the present chapter were treating social anxiety disorder (SAD), they found no difference between the two formats when the same text was used as either a book or as part of an online programme (Furmark et al. 2009). However, ICBT and bibliotherapy are not identical because while undergoing ICBT, a patient will be required to complete at least some tasks using the online programme and/or while interacting with a therapist using modern information technology (Marks et al. 2007).

Another important field with a close affiliation to the ICBT community is the field of computerised psychological treatments and programmes (Marks et al. 1998). Many of these treatments and programmes were developed before the Internet. Beating the blues, a programme for mild to moderate depression (Proudfoot et al. 2004), was recommended by the National Institute for Health and Clinical Excellence (NICE) in 2006 after several controlled trials (Marks et al. 2007). Just like bibliotherapy and ICBT, computerised CBT (sometimes referred to as CCBT often delivered on a CD-ROM) and ICBT share many similarities. The main differences are the platform needed to deliver the programme (any computer vs. a computer with Internet connection) and the way support is provided (by phone vs. by Internet/e-mail). Apart from these differences, computerised CBT and ICBT programmes can appear identical. However, literature on computerised assessments points to another slight difference between computerized CBT and ICBT (Epstein and Klinkenberg 2001): ICBT assessments are often online (such as online questionnaires), whereas, computerized CBT assessments are given offline.

There are two additional promising roles for computers in CBT. The first is through virtual reality treatment, which research has shown to be effective in exposure therapy (Côté and Bouchard 2008). Virtual reality treatment could also be integrated with ICBT, which would reduce equipment costs. This is a rapidly changing field with ongoing research on augmented reality (Baus and Bouchard 2014), and applications for conditions other than phobias and serious gaming are in the process of development (Mohr et al. 2013). The second use for computers in CBT is cognitive bias modification, which is a treatment format based on laboratory research done on various forms of selective information processing. This treatment is markedly different from regular CBT and ICBT. For example, instead of reading text and doing homework assignments to learn to avoid getting fixated on negative faces, one could do regular exercises in front of the computer (Amir et al. 2009). In most studies, these interventions were delivered in the laboratory, but in a series of studies with largely negative findings, attention training was delivered through the Internet (Carlbring et al. 2012). This is a scattered field with mixed findings, and a recent meta-analysis indicates that the effect sizes are small (Cristea et al. 2015). Some research shows that patients may benefit from attention training alone (Kuckertz et al. 2014) or as an adjunct to ICBT (Williams et al. 2013), but there are also findings that suggest there are no significant benefits of attention training (Boettcher et al. 2014a). These conflicting findings indicate that this is a field in need of more research (Boettcher et al. 2013).

In sum, computers have been used in various ways to deliver psychological treatments. According to Marks et al. (2007), computer-aided psychotherapy “refers to any computing system that aids talking treatments by using patient input to make at least some computations and treatment decisions” (Marks et al. 2007, p. 6). ICBT partly fits this description, although the focus is more on the mode of delivery than the actual use of automated decisions. In this book, we focus on ICBT, but we are aware of the blurred distinctions between ICBT and other fields. Therefore, we acknowledge the existence and contributions of related fields such as virtual reality treatment and face-to-face treatment using web cameras (Storch et al. 2011).

This leads us to the topic of this book – ICBT. The terminology is far from consistent, and various names are used and have previously been used both interchangeably to describe the same online treatments and as separate terms to describe differing treatments (Barak 2013). For example, not even the term “Internet” is used consistently to refer to Internet-based interventions. A few alternative terms are “web based”, “online therapy”, “computerized psychotherapy”, “digital interventions”, “e-therapy”, “telepsychiatry”, “cybertherapy”, “robot-based psychotherapy”, “Internet interventions” and “CCBT”. This is of course problematic when reviewing the literature, and the often very short descriptions of the interventions do not help to clarify. Here we will use the term ICBT, by which we mean an intervention that is mainly delivered using the Internet and modern information technology (e.g. accessed via different platforms) and that is based on cognitive behaviour therapy. Most of the studies and applications covered in the book will be therapist guided, but automated ICBT is mentioned.

The question of how it all began depends on who tells the story. Our presentation here is inevitably coloured by our own history, and the fact that as researchers, we based our findings on peer-reviewed publications. We acknowledge that it is possible that some clinicians have been using the treatment practices discussed in the research, but have not published any documentation (as is the case with clinicians using telecommunication software such as Skype in therapy). In research, there is also often a long lag between starting a project and finally getting the research published, which can take years. Because we know that early on multiple CBT researchers began to comment on the potential uses of the Internet (Riley and Veale 1999), it is likely that several research groups had more or less the same idea at about the same time. Researchers in Australia were early, with MoodGYM’s research on depressive symptoms (Christensen and Griffiths 2002) and the late Jeff Richards’ research on anxiety disorders (Klein and Richards 2001). Other early research took place in the Netherlands, such as research on Interapy treatment (Lange et al. 2000). These researchers began their work in the mid- to late 1990s, as did researchers in the United States such as Lee Ritterband et al. (2003). In Sweden, we were fairly early as well, with the first project starting in 1998, which subsequently lead to the first controlled trial on the treatment of headaches using the Internet (Ström et al. 2000). This was followed by a long line of research projects, but it is worth mentioning here the early work on tinnitus (ringing in the ears) that was rapidly implemented in regular health care at the audiology department in Uppsala, Sweden, by the turn of the century. This was followed by a controlled trial (Andersson et al. 2002), and the effectiveness data was published soon after (Kaldo-Sandström et al. 2004). This is probably one of the earliest implementations of ICBT in regular health care, and the treatment is still being used as a standard healthcare practice (Kaldo et al. 2013).

Early work in Sweden was also done on psychiatric conditions, beginning with panic disorder (Carlbring et al. 2001) and depression (Andersson et al. 2005b). This leads to a collaboration between clinical psychiatry and academic psychology (hence the three of us writing this chapter together), which subsequently resulted in a clinical implementation at the Stockholm health care services in Sweden known as the Internet Psychiatry Unit (internetpsychiatry.se). This work began as projects during 2002, and subsequently in 2007, the unit was opened for public use in regular health care in the Stockholm county region. There were several other implementations at about the same time in other countries around the world, but the implementation varied according to the country’s healthcare system. There are distinct differences among tax-funded healthcare systems in countries like the Netherlands, centralised units in Australia and general practices in the United Kingdom like IAPT (increasing access to psychological therapies). In general practice systems such as that of the United Kingdom, the implementation of ICBT lagged behind the research activities.

On the subject of research activities, it is remarkable that so many controlled trials have been conducted over a fairly short period of about 15 years (Andersson 2014). The number of controlled trials on ICBT for a range of conditions has greatly surpassed many other related fields, such as virtual reality, possibly bibliotherapy and some forms of psychotherapy that have been around for much longer. We suggest three reasons why ICBT has been possible to study in more than 100 controlled trials (Hedman et al. 2012). First, many time-consuming routine tasks in research, such as entering data from questionnaires, are not needed in ICBT research. Second, recruitment is publicised through advertisements (in both social media and regular newspapers). Interested participants then enter the screening phase and then are finally interviewed by a clinician. Third, in contrast to many high-quality CBT trials, a vast majority of ICBT trials in Sweden have been done using psychologist student therapists with little training apart from their basic CBT training. Interestingly, and to our initial surprise, as feedback from the guided ICBT trials, ICBT appears to benefit participants as much as face-to-face CBT.

One of us, along with help from the other two authors and a PhD student, initiated the formation of an organisation devoted to ICBT. Andersson and Lindefors invited researchers for an initial informal meeting on “Internet and CBT” in 2004 at the Karolinska University Hospital in Stockholm. During the meeting, a decision was made to form an association named the International Society for Research on Internet Interventions (ISRII) (see www.​isrii.​org). The next meeting was held in 2006 at a somewhat less fashionable hospital setting, but more people attended compared to the first meeting and it became clear that this new field was advancing rapidly, with many new trials completed just 2 years after the first meeting (Ritterband et al. 2006). Following Per Carlbring who handled a mailing list, Lee Ritterband assumed the role of managing the second mailing list and first website. The third meeting of ISRII was later held at the University of Virginia in Charlottesville, Virginia, USA, in 2007. By that time, the organisation was very active with increasing number of people on the mailing list. The fourth ISRII meeting was in Amsterdam, Netherlands, and was combined with a national meeting, but many ISRII members attended and presented separately for the ISRII community. The Amsterdam meetings were hosted by the Netherlands Institute of Mental Health and Addiction Trimbos instituut, the Vrije Universiteit of Amsterdam and the University of, in 2009. The following meeting was in Sydney, Australia, in 2011, hosted by the Australian National University. More people, over 160, attended the meetings during those warm days in Australia. It was then time to return to the United States, where the sixth meeting was held in Chicago in 2013 with more than 260 people in attendance. The most recent meeting was in 2014 in Europe, this time in Valencia, Spain, and almost 300 delegates came. The steadily increasing number of ISRII members corresponds nicely to organisation’s goal of bringing together researchers in the field of Internet interventions and other interested parties. There are now national associations for Internet intervention research and development. Additionally, there is the European Society for Research on Internet Interventions (ESRII; www.​esrii.​org), which has had three meetings already, beginning with one in Lüneburg, Germany, in 2012, followed by another meeting in Linköping, Sweden, in 2013, and the most recent meeting in Warsaw, Poland, in 2015.

In sum, ICBT is a steadily advancing field that already includes two international associations and one scientific journal. Additionally, studies on ICBT are frequently presented at conferences in psychiatry and clinical psychology, and most major journals in those fields publish ICBT studies as well. The materialisation of implementation from innovation likely would have been longer without the ISRII and the ESRII and the scientific publications.

In each of the following chapters, different conditions, programmes and outcome studies will be presented. We will not assume that all readers are familiar with ICBT, so we will now present a brief overview on how clinician-guided ICBT can be conducted. Several texts are available for a more detailed description (Andersson 2014), but the basic parts will be outlined here.

The first thing needed for ICBT is a stable and secure electronic treatment platform. The treatment platform is where the treatment is presented online using an Internet-connected device, where direct communication between a patient and the clinician takes place, and often as well as where assessments are given. Even if there are different technical solutions and regulations regarding data security issues (Bennett et al. 2010), most contemporary systems will require a secure login and often resemble Internet bank systems, e.g. used to pay online bills. In other words, systems are encrypted and often use a double authentication procedure at login. For example, a patient may need to first use a password to log in, and then a separate number via a card reader or a text message sent to their personal mobile phone (2015). For the future, we would expect that safer and more reliable systems will increase information integrity and improve user friendliness with smart individual authentication in the individual’s personal Internet device.

Most ICBT systems have questionnaire data capacities embedded in the systems. This is a separate but related topic and there are numerous studies on the psychometric properties of online administered questionnaires. Here it suffices to say that online data collection is often both efficient and reliable, but that it may be best to use the same format consistently rather than switching between online and paper-and-pencil administration (Carlbring et al. 2007).

The second necessary component of ICBT is a proper treatment programme. In almost all programmes, the main format of delivery has been text, usually in the form of book-length text materials (Andersson et al. 2008). Programmes tend to be based on face-to-face manuals and self-help books, but streamed videos and audio files are increasingly included. With high-speed Internet access, it has also been easier to use interactive programmes and even virtual reality components. Modern smartphone applications (apps) are also increasingly integrated into ICBT. The duration of the treatments mirror face-to-face CBT. For example, 10 weeks is the treatment time required for panic disorder both in face-to-face CBT and ICBT (Carlbring et al. 2006). However, ICBT programmes are sometimes shorter in research, and there are examples of programmes that have been extended in time when implemented into regular health care (El Alaoui et al. 2015). In the remaining chapters of this book, more information will be provided on the actual components included for the particular mental illness (e.g. major depression), but here we can note that clinician-guided ICBT includes homework assignments, followed by feedback and encouragement on the assignments delivered through the securely closed platform. In addition to feedback, the clinician can answer questions on the programme, but most correspondence usually takes the form of support and advice on how to complete the assignments and progress through the treatment (Andersson 2014).

There are conflicting results on the third aspect of guided ICBT, clinician guidance. Several systematic reviews have found that guidance reduces dropout and probably increase the effects (Baumeister et al. 2014), but it is possible that the need for a therapist differs depending on the conditions. It may also be true that access to a clinician when the patient asks for it and automated reminders are enough for some patients (Titov et al. 2013). However, some studies clearly support the superiority of clinician guidance (Kleiboer et al. 2015). Additionally, although it is not trivial how the correspondence with the patient is handled, it is probably the case that for a vast majority of patients it does not really matter who the person is giving them guidance (Almlöv et al. 2011; Almlöv et al. 2009). Moreover, the support can be mainly technical and practical in nature and does not need to be psychotherapeutically oriented (Titov et al. 2010). On the other hand, there may be therapist behaviours that are especially helpful for some patient groups, like a forgiving attitude towards non-completion of homework in the treatment of generalised anxiety disorder (Paxling et al. 2013).

As with any treatment for psychiatric disorders and in health care in general, ICBT raises questions regarding ethics and possible side effects. Although there are general aspects to consider such as human rights and the principle of doing no harm, there are also some considerations that are unique to ICBT. For example, whether it is acceptable to provide psychological treatment from a distance using the Internet may differ between countries like Sweden and Norway. In the United States, regulations may vary depending on which state the practitioner resides in. There is an emerging albeit small literature on the ethical aspects of ICBT (Dever Fitzgerald et al. 2010). The current consensus is that ICBT should be subject to the same ethical regulations as face-to-face therapy, but there are some additional aspects to consider in ICBT such as data protection/security, clinician responsibility, the possibility of anonymous treatment, etc.

One important aspect of ICBT is the possibility that adverse events will follow the intervention and that there is a clear link between those events/symptoms and the treatment. The literature on this topic is growing and recommendations to report negative outcomes in trials have been published (Rozental et al. 2014). There are also empirical studies on the topic. Data from four large clinical trials (total N = 558) revealed that 9.3 % of patients reported some type of negative effects (Rozental et al. 2015a). Another study from the same group reported that 14 % experienced negative effects in a trial on social anxiety disorder (SAD). These “negative effects” were defined as unwanted negative events that the patients related to the treatment (Boettcher et al. 2014b). Even though severely negative outcomes following ICBT are fairly uncommon, they should be reported, which has been the case in recent studies.

In this book most, but not all, psychiatric conditions for which ICBT has been developed are included. Here we will discuss some of the conditions that have been researched using controlled studies, as well as comment on the large literature on somatic conditions and studies on subclinical problems like perfectionism.

There are a few studies that have focused on guided ICBT for the treatment of specific phobias (Andersson 2014) and some studies that do not focus on, but include, patients with specific phobias, such as research done on the Internet version of the programme FearFighter (Schneider et al. 2005). Another example is a fairly large (N = 212) recent trial done in the Netherlands. There was a high rate of attrition, and the results of the trial showed a small effect (I am not 100 % sure this is what you mean by “small effects”), which is noteworthy because these results differ from those of many other trials in the field (Kok et al. 2014). This study did not only include specific phobia patients; other patients were included as well. To our knowledge, only two controlled trials exclusively done on adults with specific phobias exist. These two small studies from Sweden compared live one-session exposure to guided ICBT (Andersson et al. 2009, 2013). In the first trial, 30 participants with spider phobias were included, and the results showed that both the one-session treatment and the ICBT condition were effective, but the one-session live treatment condition participants did better on a behavioural approach test (BAT). Following treatment, evaluation revealed that 46.2 % of the ICBT group and 85.7 % of the live-exposure group achieved clinically significant improvement. In the second trial (Andersson et al. 2013), 30 participants with snake phobias were included. Results were similar with no difference in self-reported outcomes, but again there was a difference in BAT scores. The post-treatment evaluation revealed that 61.5 % of the ICBT group and 84.6 % of the one-session group achieved a clinically significant improvement. In a Cochrane review on guided ICBT for anxiety disorder, only one trial could be included (Olthuis et al. 2015), and the 2009 trial from Sweden was selected. One uncontrolled trial on guided ICBT for children with specific phobias has also been published (Vigerland et al. 2013) and will be covered in a separate chapter in this book. Thus, there are only a few small studies on guided ICBT for the treatment of specific phobias, so there is a need for larger replications of the Swedish findings. It is possible that online versions of virtual reality treatments or augmented reality using smartphones will be the next step, rather than the self-guided exposure treatment that has been the focus of the previous ICBT trials on specific phobias.