Barriers and facilitators for the implementation of blended psychotherapy for depression: A qualitative pilot study of therapists' perspective

1Depression is the most common mental health disorder with an estimated lifetime prevalence of 12.8%. Psychological treatments have shown to be effective. However, the majority of individuals with depression remain untreated. Reasons for the low uptake of psychotherapy include: (a) referral rates by general practitioners to psychotherapy are often low and vary between 21% to 58%; (b) patients are confronted with long waiting lists, especially in rural areas; (c) limited time resources and availability of licensed psychotherapists and; (d) attitudinal barriers often hinder patients from utilizing necessary treatments.

2Using internet- and mobile-based interventions (IMIs) to deliver psychological interventions could be a solution to overcome existing problems. IMIs have several advantages, e.g. an extensive coverage due to independence of time and space as well as therapy for patients at their own pace. There is convincing evidence that such approaches are effective for depression, show higher effects with increased guidance and even similar effects, when head to head compared with face-to-face (f2f) treatments.

3However, stand-alone IMIs might not be acceptable and suitable for all patients. Blended approaches use “one integrated standardized CBT-treatment protocol combining face-to-face and digital components to the best clinical benefit for patients and therapists” (Riper 2017). Possible advantages of such approaches include (a) acceptability to patients for whom pure online-based interventions might not be a suitable treatment option, (b) improved treatment adherence rates compared to purely online guided interventions, (c) the potential of time saving through processing psycho-educative content online and using therapy sessions to intensify therapy subjects or by treating more patients.

4Blended therapy (BT) has not yet been well studied and its implementation is still scarce. The uptake of eMental health by routine practice is so far limited. An earlier RCT compared BT for depression – f2f sessions and a smartphone application - against a full behavioral treatment and found significant improvements in both groups, but could not establish whether the BT was non-inferior to a full BA treatment. A recent study showed that BT for depression was more effective than psychotherapy alone and could be a promising option to consider in future treatment for depression. The usage of blended therapy depends on its clinical effectiveness and on its acceptance by key health care professionals. An understanding of relevant barriers and facilitators for successful implementation from the perspective of relevant stakeholders is considered to be essential. Therapists are one such important stakeholder.

5Barriers and facilitators for the implementation of BT from therapists' perspective are sparsely researched. Studies indicated that the use of a blended approach in therapy has been low, with 45% of psychologists confirming use of email and mobile text messages but having a neutral attitude towards it, and with 12% of therapists confirming experience with IMIs. However, a promising 90% of therapists have signaled their readiness to use it in the future, especially as a supplement rather than an alternative to f2f therapy. To blend IMIs with f2f therapy, therapists have asked for more information as well as training to enhance usage skills. One of the few studies that examined attitudes towards BT for depression among care providers, found that BT was more acceptable as an effective treatment for depression for all severity levels, compared to stand-alone IMIs. However, still only 55% of care providers perceived BT as suitable for moderate depression, and 22% found it suitable for more severe cases. In a Delphi study potentially suitable blended protocols (content, sequence and ratio) for depression treatment were examined by 12 therapists. The strengthening of patients' self-management was considered an important facilitator. Further key findings were that the possibility of tailoring the treatment to fit the individual patients' needs was perceived as essential, as was the possibility of adjusting the amount and ratio of online modules and f2f sessions according to the patients' problems, skills and characteristics. A following initial pilot evaluation observed that therapists evaluated BT as a helpful tool in providing evidence-based treatment in secondary care.

6A recent scoping review on determinants for a successful implementation of stand-alone IMIs for depression in regular care showed that guided support from health professionals emerged as a relevant theme. The guided support was mostly performed by therapists who aimed to enhance treatment adherence to the intervention. Important barriers that were identified included limited knowledge about and experience with these approaches, the lack of availability in routine mental health care, and a lack of training and supervision that would allow care providers to get acquainted with the treatment format. In a recent survey 88 care providers reported a moderate level of knowledge about IMIs. They also expressed concerns about the limited internet literacy of some patients, concerns regarding online safety, and concerns about the potential negative effects on the therapeutic alliance, as well as on patient commitment as important barriers to implementation. Reduced cost of treatment and a better access to treatment compared to TAU were mentioned as main facilitators for an integration of stand-alone IMIs in the health care system. A qualitative study with five interviewed therapists showed, that they felt hindered by the existing poor integration of a guided IMI for depression in primary care as well as the difficulties of incorporating IMIs in their work routine. To enable them to use BT, therapists asked for a more flexibility in delivery of the treatment as well as the integration of f2f sessions within a blended approach.

7With regard to stand-alone IMIs for the treatment of mental health disorders a literature review suggested an accurate diagnosis, a user-friendly layout, a comprehensive treatment (not overly technically advanced), and a provision for guidance by a therapist as crucial determinants for a successful implementation. Important hindering factors that have been identified for eHealth solutions in general include problems with interoperability between systems, a lack of adaptability, too little resources for familiarization with technology, poor motivation of professionals, a lack of communication about advantages as well as safety concerns, a lack of confidence, and a general rejection of telecare and low operating skills.

8Despite the evidence-base for effective interventions, their transfer into routine practice is often not achieved. To bridge the gap between research and practice, the use of a theoretical background regarding behaviour change in health professionals is recommended, as it guides the selection of factors that should be examined. One example for implementation theories is the Normalization Process Theory (NPT), which includes factors that facilitate and prevent the routine incorporation of complex interventions into daily practice and aims to normalize their application. NPT also includes the broader organizational context, social norms and group processes, but so far NPT has been less applied in research. The ten-stages model for planning change belongs to the process models, and it aims to describe and guide the process of translating theories and research results into practice when implementing a new intervention. The Theoretical Domains Framework (TDF) aims to identify domains that influence the implementation of interventions. The TDF consists of 14 domains of potential behavioral determinants that were developed by the consensus of experts and grounded in psychological theories and that have been validated. It enables researchers to explain barriers and facilitators of behavior change; thus supporting the implementation of evidence-based practice. It was used in several studies for the development of qualitative and quantitative measurements. In the current German treatment context, therapists treat their patients only with f2f therapy and do not incorporate online modules. However, blending f2f sessions with IMIs does lead to a behavior change; so the identification of influencing determinants of the therapist's behavior by the use of the TDF is crucial.

9Past studies examined the use of IMIs in general in different target groups. Previous studies with therapists used mainly quantitative surveys, had no link to theories of behavior change, and asked therapists, most of whom had no user experience, to report their attitude, acceptance, and intention to treat with IMIs. One qualitative study with focus groups aimed to identify the optimal usage of blended treatment protocols for depression; but therapists differed in their user experience and assessed BT protocols that were developed on the basis of a prior quantitative survey. A following study conducted an initial evaluation of this developed BT in specialized mental health care. This study adds to the literature several unique features: The study first examined the therapists' perspective on barriers and facilitators for the implementation of BT for depression. The study employed a qualitative theory-based approach to gain in-depth insights and addressed different domains of behavior change by interviewing the therapists after they had treated patients with BT.

10The aim of this exploratory pilot study was to generate hypotheses and identify facilitators and barriers for implementing a 13-weeks BT based upon the perceptions of trained psychotherapists who experienced the application of BT in the treatment of depression while participating in the German E-COMPARED trial.

11The current research project was part of the German study arm of the European Project E-COMPARED (www.e-compared.eu), that was the first multi-center study which aims to evaluate the clinical and cost-effectiveness of a blended depression treatment across eight European countries. The main objective in Germany was to evaluate the effectiveness of blended Cognitive Behavioral Therapy (bCBT) for adults with a diagnosis of Major Depressive Disorder (MDD) compared to treatment as usual (TAU) in primary care. BT was conducted by psychologists in a publically funded outpatient university clinic. TAU was the routine care patients receive by GP, the most common treatment for depression in Germany. Patients were recruited through local general practices from February 2015 to August 2016 and randomly assigned to the treatment conditions TAU (n = 87) vs. bCBT (n = 86).

12The study was approved by the ethics committee of the German Society of Psychology (DGPs) and registered in the German Clinical Trials Register (DRKS00006866, on 2 December 2014). All therapists provided informed consent.

13The blended approach is a new treatment format in Germany and in other countries. It has an innovative character regarding the inclusion of web- and app-based components and ecological momentary assessment. BT for depression consisted of six therapy modules on an online platform (Moodbuster), a mobile app, as well as six f2f sessions with a therapist. It was delivered within 10 to 13 weeks by psychologists as an alternative treatment for general practitioners' care. This duration of treatment is considered as short time intervention in the German health care system. The platform provided access to treatment modules, homework exercises, mood graphs, a calendar, and a messaging system. A therapist portal enabled therapists to monitor their patients' treatment course, provide feedback on exercises and progress or write reminder or motivation emails to enhance adherence. The online modules based on established CBT-principles for MDD and consisted of 10 lessons for psychoeducation, behavioral activation, cognitive restructuring, problem solving, sports and exercises, relapse prevention. After editing two mandatory online lessons, patients had access to all of the remaining modules except the relapse prevention module. That module was released only after finishing their last f2f session. One module was optional. The patients were instructed to work on one module lesson at a time and to complete it in one week. Together with his or her therapist, each patient decided upon the sequence in which to complete the modules. Patients worked on the online lessons weekly at home, used the app for daily ecological momentary assessments (mood state, activities, sleep habits), and got automated tailored reminders and individualized motivational messages. On-site f2f therapy started with a technical introductory, lasted 50 min every two weeks, and enabled therapists to refer back and forward to web-based modules and mobile-based ratings in order to consolidate themes and structure the treatment.

14The patients treated by the interviewed therapists of this study (N = 71) were on average 43.4 (SD = 13.16, Range = 19 to 70) years old, mostly female (59.2%), highly educated (53.5%), and had a paid job (71.8%). 71.8% had no earlier experience with psychotherapy. 90% of patients completed all six f2f sessions, 75% all ten online sessions. They had 5.7 f2f sessions on average (SD = 1.06, Range: 1 to 6) that lasted 57.3 (SD = 5.02, Range = 48 to 77) minutes.

15Eight therapists treated patients in the study and were invited to join expert interviews. In total, five therapists took part in semi-structured interviews. Three could not participate because of interference with current occupation, as they no longer worked for the trial center.

16All participants were women, young (M = 28.40 years old, SD = 2.52), held university degree in psychology (master, diploma) and were undergoing further training as advanced clinical CBT-therapists. Their average experience as therapist was 24 months (SD = 29.39), but without previous experience in BT (see Table 1). They were employed between May 2015 and December 2016 and treated on average 16 patients with BT for depression. All the therapists were trained in the BT protocol under the supervision of an experienced, licensed psychotherapist.

17We developed a qualitative design method with a theory-based approach to gain insights into the experiences of the therapists who worked with BT. This method is recommended as the best approach to identify barriers and facilitators and to bridge the gap between research and practice.

18The interview guide was developed after a literature review according to the theoretical domains framework (TDF) with 14 domains of potential behavioral determinants and scholarly exchange with clinical experts. The semi-structured interview guide (see Table 2) should elicit therapists' perceptions of barriers and facilitators while applying BT. When needed, guided prompts enabled gathering more information.

19All semi-structured interviews were conducted by the first author (IT) at the workplace of the interviewees between April and November 2016. The average duration of an interview was 99 min (SD = 12.10). Interviews were audio-recorded and then transcribed by the second author (KS) verbatim based on a transcription guide. Anonymity was assured by using code numbers instead of names.

20A qualitative content analysis was conducted, drawing on inductive and deductive approaches and using standardized methodical steps in qualitative research.

21Following the inductive approach, codes were developed from the raw data, based on themes relevant to addressing the research question. First, a list of codes based on 60% of material was developed after identifying emerging (sub-) themes and discussing excerpts, the organization of the categories, and the possible meanings of text frames within consensus meetings (KS, IT). Second, the entire interview material was read, labels were attached to text parts related to the list of codes, and new emerging themes were added to the list of codes. During this coding process, new categories emerged and old ones were changed in order to fit the data. A sentence or paragraph could be coded as containing aspects from one or more categories. A preliminary categorization guide based upon 100% of the material was developed and discussed in consensus meetings. The consensus meetings reached final agreement on code definitions and excerpts, the structure of the code system, a final list of codes, and coding rules. The clinical expertise of a CBT therapist (IT) enriched the study's data interpretation by discussing the possible meanings of text frames in the context of applied CBT. Third, two coders (KS, SH) independently coded all transcripts of the interviews with the five therapists in accordance with the final code list. There was sufficiently high agreement between the coders with Cohen's Kappa k =.85. Forth, data saturation (i.e., each category being mentioned by at least two therapists) was not expected for the small sample size of this exploratory study. However, 97% of the categories were mentioned by at least two or more therapists, while only two of the categories were mentioned by only one therapist. This level of data saturation is valid for case analysis. Fifth, to ensure the validity of the identified categories, the five therapists were questioned as to their individual agreement with the resulting barriers and facilitators as indicated by data analysis. The agreement rate per individual therapist was M = 90% (min = 78%, max = 100%). In total, all five of the therapists agreed with 66% of the themes, while four out of five therapists agreed with 84% of the themes. Thus, the identified barriers and facilitators yielded a good validation result with an agreement rate of at least 40% to 100%. Sixth, key themes were defined as emerging themes mentioned by all five therapists. These key themes are reported in the results section. All categories, including other themes mentioned by four or less therapists, are listed in tables and figures.

22Following the deductive approach, coding of the whole material was done by referring emerging themes to the domains/constructs of the TDF and classifying the reported experiences within each domain as barrier and facilitator. The independent coding by two persons yielded a mild to moderate agreement with Cohen's Kappa κ =.68.

23All qualitative analyses were conducted with the software tool MAXQDA. Consolidated criteria for reporting (COREQ) were followed (see Supplementary Table 1).

24In sum, 29 barriers and 33 facilitators were identified. Each barrier and facilitator was categorized and abstracted into one of four main areas, as follows: (a) hindering and enabling factors that can be useful to develop an implementation strategy that embeds a 13-week BT to treat depression in Germany's primary care system, while considering the existing framework, were classified under the topic ‘implementation in the health care system’ (see Table 3); (b) hindering and enabling factors that influence the therapy process and the interaction between therapist and patient were labeled as ‘therapeutic factors’ (Table 4); (c) hindering and enabling factors that personally influenced the therapists‘ behaviour, cognition, and feelings were clustered under the label ‘therapist factors’ (Table 5); (d) while, likewise, hindering and enabling factors that personally influenced the patients' behavior, cognition, and feelings were clustered under the label ‘patient factors' (Table 6).

25All barriers and facilitators are described alongside a definition and supporting quotation. In total, the most facilitators (n = 15) were found on the level of ‘Implementation in the health care system’, whereas the most barriers (each n = 8) could be identified on the level of ‘therapeutic factors’ and ‘patient factors’.

26In total, therapists perceived more barriers regarding BT on the level of ‘patient factors’ and ‘therapeutic factors’ compared to ‘implementation in the health care system’ and ‘therapist factors’. 24% (n = 7) of the identified 29 barriers were mentioned by all therapists, and 48% (n = 14) by at least four of the five therapists. Below, key barriers mentioned by all five therapists are described with a quotation illustrating the therapists' view. All categories and subthemes and the number of therapists who mentioned the respective barrier are listed in Fig. 1.

27In this category none of the barriers were reported by all of the therapists, therefore barriers that were mentioned by four therapists are described below.

28Four therapists mentioned the immature technology and other technical problems, such as a slow internet connection or faulty interfaces, as important hindering factors. Furthermore, they noted that the therapist should have some additional resources to provide BT, such as access to the platform, the possibility of checking online activities, the capability to write and send reminders and motivational emails to their patients, and training sufficient to enable the therapist to solve minor technical problems. The lack of a concept to fund these additional activities was perceived as a barrier. But also the lack of a concept by which to embed BT into the health care system was described as main barrier, because its currently not part of routine care.

29“Although, we'll need to see how it will actually work: How would patients use it; what structures should be given? Would it be linked to a therapeutic private practice or included in outpatient mental health clinics or something? I'm a little skeptical; what will it look like actually.” (T011).

30All therapists criticized the ‘one size fits all` approach, the standardized treatment procedure for depression, and the limited possibilities to tailor the treatment to the patients’ individual needs.

31“Another partly difficult thing was that the modules were generalized. It was hardly possible to give a module about activity to one and a module about ‘saying no’ to another patient. However, it was already obvious that the ‘one size fits all’ approach is not working everywhere.” (T007).

32Consistent with this perception, all the therapists wished to be autonomous in deciding how to use online lessons in their patients' therapeutic treatment. The therapists wanted to choose the number and ratio of applied f2f and online sessions, to choose the content, and to decide the sequence of online modules and which patients they treat with BT. The therapists recommended 11 (min = 7, max = 25) f2f sessions on average for patients in the short time blended approach in Germany. All wanted more flexibility and no pre-described protocol regarding the ratio of online and f2f sessions, as the following examples show:

33“A certain kind of flexibility. For example with the number of f2f sessions. So no one should say that with using this therapy there are only 6 f2f sessions possible, but more the possibility to decrease or increase this number.” (T010).

34All therapists further mentioned that the low and limited number of f2f sessions for each patient reduced the potential benefit of the approach. For example, the time was not sufficient to help patients to stabilize achieved therapeutic gains.

35“Moreover, for 90% of all my patients six sessions were too few.” (T004).

36Moreover, permanent or unresolvable technical problems caused anger, frustration, and demotivation in both patients and therapists. The therapists all agreed that these barriers and providing technical introduction in the first therapy session hindered the establishment of the therapeutic alliance.

37There was only one barrier stated by all therapists. They perceived the need to remind patients to complete online-lessons as frustrating, onerous, and time-consuming, as one therapist outlined:

38“Reminding people takes most of the time. ‘Do this, do that’ (...). It is a little exhausting and sometimes frustrating.” (T010).

39Regarding important factors on the patient level, all therapists assumed that those patients having reservations towards IMIs may not be suitable for this treatment approach and should therefore have the possibility to choose stand-alone f2f treatment. All therapists reported the impression that some patients perceived BT as a temporary intervention to bridge the time until their start of a traditional f2f therapy, others perceived online components as less important than f2f therapy.

40“In the end, I think some of them did not think of those exercises as an important part of the intervention.” (T011).

41In addition, they pointed out potential disease-related contraindications for the use of BT, such as a high disease burden, chronic course, or suicidal tendencies.

42“I think it is difficult for example with chronical depression or the latest of many recurrent depressive episodes. Especially because of the short time.” (T008).

43Therapists perceived more facilitating then hindering factors in the 13-weeks BT for therapist factors and implementation factors, and more hindering factors on the level of patients and implementation in the health care system. 48% (n = 16) of the identified 33 facilitators were stated by all therapists, 76% (n = 25) by at least four. Below, key facilitators that were mentioned by all i the therapists are described with a quotation illustrating the therapists' view. All categories and subthemes and the number of therapists mentioning the respective facilitator are listed in Fig. 2.

44All therapists judged BT as a modern, contemporary and adequate treatment approach, since nearly everyone is familiar with using the internet and has daily access to it. The content of the modules was assessed as therapeutically meaningful as well as relevant for MDD patients, as a therapist expressed below:

45“So what really appealed to me the most was this very nicely designed website with the really well-prepared lessons, which provided information in a nice way and gave examples and exercises.” (T011).

46Other major facilitators stated by all the therapists were that BT can require fewer resources as compared to traditional psychotherapy, and that allows therapists to treat more patients over the same period of time. They indicated also that BT offers the possibility to provide more immediate access to treatment and thereby eventually reduces the treatment gap in mental health care.

47“In the end it is possible to provide more therapy. Since BT is also something young therapists and therapists in training can easily work with, certainly more therapies could be provided.” (T011).

48Therapists judged blended care as an attractive approach for hard to reach patient groups, e.g. men, young patients, people with reservations concerning traditional psychotherapy who might favor using IMIs, and those who might prefer fewer f2f sessions due to conflicts, for example, with other obligations in their daily lives.

49All therapists stated the necessity of offering educational and training sessions such as workshops and test-accounts in order to familiarize therapists with the BT approach. In addition, promotional activities could help to enhance the prominence and acceptability of BT. The national press could be enlisted to disseminate information about BT; and investments in advertising and educational activities could be undertaken on a national level.

50“GP's or medical specialists need to have the blended care approach in mind to be able to propose and recommend it to patients.” (T008).

51Furthermore, therapists also appreciated the intuitive usability, design, and logical structure of the online platform and mentioned usability aspects as crucial for implementation targets.

52As one important therapeutic factor, all therapists indicated that the online components facilitated the preparation of f2f sessions for both therapists and patients. Patients having access to information between f2f sessions and after finishing treatment was seen as a big advantage, as this facilitated the therapeutic process and allowed patients to repeat lessons on their own and to reflect on their behaviour without having a therapist present. But therapists also liked the possibility of monitoring the treatment course and assessments online and referring to it in f2f sessions. They also noted that the online treatment modules helped them to structure the treatment, which might be beneficial for the overall treatment outcome.

53“Patients can revise what we talked about, get access to all exercises and add something later on. Furthermore, they can repeat some modules after the official end of therapy.” (T007).

54In addition, therapists expressed, that this blended 13-weeks intervention effectively helped to reduce symptoms of depression in a short time frame, as evident in the following sentence:

55“As patients came into treatment everything seemed quite bad, but there was this huge benefit of BT in such a short time.” (T011).

56All the therapists noted that their time was saved due to the availability of therapy material and pre-gathered information online. Having access to a digital tool kit supported their therapeutic work, as they used a pool of information and exercises with a lot of input to focus on specific issues.

57„It is helpful to know that you have this repertoire of exercises and you can focus on one or the other with every client. Maybe there are some modules that are less useful for some of them and that is also okay.” (T010).

58All therapists felt enabled to perform BT by being aware of the role and tasks of a BT therapist. They reported that they felt no identity conflict with their role, and all were confident that their knowledge of BT was sufficient. They were especially confident that they could integrate the web- and app-based parts into f2f sessions. The therapists observed, as well, that being knowledgeable about the functionalities of the website were important facilitators.

59The therapists expressed the following personal traits as essential facilitators for a BT therapist: an empathetic attitude; a high tolerance for frustration; persistency; good writing skills; and the ability to build up a therapeutic alliance with one's patient in a short time frame. Beneficial therapeutic skills were outlined as exemplified below:

60“As a therapist, I also think you should have a certain enthusiasm about online treatments. Not only as a skill, but as a requirement. I need to have some kind of ability to teach and motivate people, why it's worth completing an online lesson additionally.”(T007).

61The therapists indicated that most of their patients expressed great interest and willingness to participate in BT. The therapists observed that their patients felt motivated by this innovative and exciting approach and wanted to engage more in some treatment interventions, such as mood and activities ratings on the app, and using the online calendar in the online portal, in comparison to stand-alone f2f therapy.

62“NONE [patient] would have EVER drawn so many mood curves with a pen. But there is a high intensity of drawing those in the app. I think this is useful even after the end of the project [because they continue with mood ratings].” (T007).

63With reference to the TDF three domains that influence behaviour were identified as barriers, and eleven domains as facilitators. Concerning facilitators, therapists reported they had sufficient knowledge and skills to perform blended care, experienced no identity conflict, and judged blending online modules with f2f therapy as part of their role. They were optimistic according to a positive course and outcome, expressed a high intention and motivation to use it, and found it easy to draw attention to it. Furthermore, therapists expected a positive social influence and felt supported by their positive emotions and behavioral regulation mechanisms while performing BT. Regarding barriers, therapists reported rather reduced self-efficacy because applying BT was also perceived as stressful. Furthermore, they mentioned some hindering environmental context factors (e.g. no existing reimbursement concept) for implementing BT. They also reported their uncertainty about the cost–benefit ratio of BT.

64The current study identified important facilitators and barriers for the use and implementation of a 13-weeks BT for depression from therapists' perspective.

65The barriers that have been identified by most therapists as crucial for a successful implementation in the health care system were similar to the barriers in previous studies on non-blended, guided stand-alone IMIs and, in general, other e-health solutions; namely, concerns about data safety and the absence of a clear concept as to how BT can be embedded in the health care system. Especially important is the lack of funding solutions for the additional effort associated with a blended approach; namely, monitoring online activities, sending emails to remind and motivate their patients to complete the online modules, writing feedback messages to exercises or providing technical support. A partial solution for the necessary additional efforts noted by the therapists (aside from paying the therapist for all of his/her extra work) might be the inclusion of other professionals, such as mental health nurses and/or psychologists, to do the tasks that don't require therapeutic expertise. All therapists identified the repeatedly occurrence of technical problems as barrier to use BT. These technical problems resulted in anger and frustration and distracted the therapy process. Technical problems as hindering factor or the need of user-friendly technical solutions has been repeatedly reported in the literature generated from studies of the use of IMIs for the treatment of depression, e-health solutions in general, and non-blended IMIs. The successful implementation of BT will certainly require a sophisticated technical solution that is free of typical teething problems.

66On the other hand the implementation of BT could be facilitated by other factors that have been also found in studies on stand-alone IMIs, e.g. that IMIs were widely accessible to patients and could help to overcome the treatment gap in the health care system. If the 13-weeks BT is conceptualized to save therapists time, more patients could potentially be treated with the same number of trained clinicians. However, a recent study showed that offering therapist the option to use online modules to deliver some aspects of treatment online, does not automatically lead to lower costs, as therapists simply provided the online treatment modules on top of the f2f sessions. Such a procedure would not reduce the treatment gap. Hence, future research needs to investigate strategies that would motivate therapists to reduce the number of f2f sessions per patient because reducing clinician time per patient is a goal of the 13-weeks BT in Germany. Other important facilitators for the implementation of BT included the offering of training to therapists to acquaint them with the new treatment format and the focusing on dissemination activities designed to enhance the acceptability of BT among therapists and patients. Results showed that several barriers and facilitators for BT are in line with results of previous studies of IMIs and e-Health solutions. If all technology-based solutions for health care would develop together an implementation strategy, profitable synergy effects could emerge.

67Furthermore, the use of BT will also be influenced by its possible consequences on the treatment process. The therapists identified several factors that would hinder their therapeutic work: all therapists demanded more autonomy regarding their decisions about how to use BT and more customizability of the online modules. They rejected the ‘one size fits all’ approach. They objected to the blended approach's lack of individualisation. The therapists want the authority to determine the number and ratio of f2f and online sessions for each of their patients. These therapists' objections are in line with the results of a recent Delphi study in which therapists were consulted about the optimal format of blended treatments for depression as well as a mixed-method approach with therapists about barriers for implementation of guided IMIs for depression. Regarding the therapeutic alliance that is burdened by technical problems, future implementation plans should be wary of providing the technical introduction in the first f2f session. Further, an IT support is needed, independent of the therapists, to address any permanent technical problems that might occur in the IMIs. The therapists also demanded the right to choose and to decide autonomously which of their patients to treat with BT. All the therapists noted several disease-related contraindications, such as severe depression, for instance, that were deemed unsuitable for such a treatment. In line with this finding, the reported acceptability of IMIs for depression was also low in a recent study, if there were contraindications. Thus, it might be important to consider BT is not suitable for specific patient groups and to inform therapists systematically about the evidence on moderating factors. Providing case reports, especially for patient groups for which the evidence is available, could be helpful in informing therapists about indications and examples of treatment. The therapists also mentioned that they need additional online content that allow for the treatment of comorbid mental health disorders. This was also recommended in a previous study regarding the implementation of stand-alone IMIs.

68On the other hand, there are important facilitators regarding the therapy, e.g. most therapists appreciated the online modules as a useful complement of psychotherapy. Thus, they prefer a blended approach, if the f2f sessions are a mandatory and integral portion of BT, while the online modules are complementary. This was in line with a recent study that showed primary care therapists wished to integrate and blend IMIs for depression and f2f therapies because they judged a blended approach as more efficient. Another study found that BT was more accepted than stand-alone IMIs, because the lack of personal contact between patient and therapist was seen as a crucial barrier in the implementation of stand-alone IMIs to treat depression. In the current study all the therapists also reported some advantages for patients; e.g. that the use of online modules could lead to empowered self-management skills and self-efficacy, which has also been reported as one of the key advantages of BT in one of the few qualitative studies on the therapists' perspective on BT. This qualitative pilot study yielded new insights from the therapists as to how the blending of f2f and online sessions can facilitate the treatment. These insights were not reported in previous studies. The therapy profits from: (a) patients' access to information between sessions and after the end of treatment modules, which leads to (b) time savings in the f2f sessions, thus enabling therapists to use the additional time to intensify the therapy; (c) the therapists' access to a digital therapy tool kit; (d) therapists' monitoring the online treatment course and use of the assessments, e.g. daily mood ratings, to optimize therapy; and (e) the preset structure of the online platform that guides the treatment course for both the therapists and the patients; while (f) providing to patients a contemporary and modern concept of treatment that fits the patients' needs.

69Interestingly, the interviewed therapists reported to have a positive attitude towards this innovative care form, but expected their colleagues to be skeptical of this new approach. Previous studies identified negative attitudes towards e-health solutions in general as barriers. But there is also evidence that therapists have an interest and a positive attitude towards the implementation of IMIs in general and BT in particular, and that is considered to be a facilitator. The therapists in this study mentioned that, after working with BT, they were ready and willing to use it in the future. This seems to be in line with the findings of previous studies where therapists' lack of experience was noted as a hindering factor to the implementation of IMIs for depression. So, enabling therapists to experience BT through workshops, through its inclusion in the curriculum of universities and/or training institutes, and through providing free educational test accounts to therapists could promote its usage. While more exposure to BT could potentially reduce the therapists' misperceptions and reservations, probably therapists would be receptive of BT if they understood that their patients are interested in this innovative treatment and are willing and motivated to participate in BT.

70The current study has some limitations. The sample consisted of only five therapists, all female. Furthermore, they worked within a structured, standardized treatment approach in the context of a randomized controlled study. In addition, the interviewed therapists had treated a limited number of patients, their average experience differed a lot and experienced therapists might express other views. Hence, generalizability and representativeness cannot be assumed. Furthermore, a wide variability of technical problems might have influenced the therapists' views. Moreover, the interviewed therapists applied for the position to treat patients within a BT approach; hence, results might be biased towards a more positive attitude of BT.

71The study's strengths included (a) a theory-based semi-structured interview guide; (b) data collection and analysis by different persons; (c) continuous consensus approach and development of a code list following guidelines of qualitative research; (d) independent coding and high interrater reliability; and e) validation of results by interviewed therapists. Certainly future studies need to explore barriers and facilitators using qualitative and quantitative measurements in larger samples, male therapists or other health care professionals, and therapists that have used BT in the context of routine care, without restrictions with regard to standardization of a randomized trial.

72BT is a new care form in Germany and most therapists are only experienced with performing f2f therapy and not blending it with online sessions in 13 weeks. Therefore, implementation approaches have to consider a wide range of determinants that are crucial for behavior change of therapists. In general, the transfer of evidence into routine practice is often not achieved, but behaviour change of therapists could be facilitated by an understanding of influencing barriers and facilitators and using it for a successful implementation. This qualitative study has shown possible facilitators and barriers for the use and implementation of BT for depression from the perspective of five therapists from which implementation efforts can learn. They don't claim to be representative, rather case-descriptive, but deliver in-depth insights and enable possible recommendations to overcome barriers for BT implementation. In addition, future research is needed on some aspects of what therapists suggested, e.g. the flexibility of use of f2f and online modules.

73Summarizing our in-depth insights in perceived barriers and facilitators by therapists, we would redesign the BT process for the German health care system in the following manner: Therapists should be able to individualize the treatment and can autonomously decide about the ratio and number of used online modules and f2f sessions. They could choose which patients they treat with BT and could consider disease-related contraindications in their decision. They should have access to online content, e.g. a digital therapy tool kit for comorbid mental health disorders in order to tailor their treatment plan. To reduce the burden of technical issues to the therapeutic alliance, the technical introduction session will follow the first f2f contact and IT support will be provided by IT specialists. The aspects of the online platform could be improved to enhance the usability (e.g. smartphone-based calendar). To reduce time pressure and additional organizational effort, e.g. for monitoring online activities, some of the therapeutic tasks could be delegated. Regarding the future implementation of BT there is a need to develop and provide trainings for therapists. In addition, a concept of embedding BT in the health care system and funding the additional effort has to be developed by involving relevant stakeholders (e.g. health insurances, professional associations, Federal Joint Committee). To prevent frustration and difficulties for patients and therapists when using the online platform, a sophisticated technical solution has to be implemented through cooperation with IT experts. In general, BT could be offered for the treatment of depression and comorbid disorders, but also for relapse prevention.

74Another interesting approach to guide implementation efforts is the ‚Fit for Blended Care’ instrument, which aims to facilitate the setting up of a personalized blended treatment by the use of a checklist for barriers and facilitators. All findings from previous, current, or future studies could be summarized and adapted to country-specific settings and be presented as a tool kit to enable professionals to apply BT. The EU project ImpleMentAll (www.implementall.eu) is a current example for such a research approach that aims to provide evidence-based answers to the problems of implementation through the development, application, and evaluation of tailored implementation strategies.

75The current study showed that a theory approach with the TDF is also useful for the intervention context of BT applied by therapists. Questions informed by TDF generated 29 barriers and 33 facilitators for what we hypothetically assume to be determinants of behavior change. Each TDF domain was classified as being perceived as a barrier or a facilitator by the therapists. We are now able to use the COM-B model in future implementation efforts and studies. The COM-B model is linked to the TDF and postulates that behavior can be influenced by the components capability, opportunity, and motivation. The guide ‘Behavior change wheel’ will provide recommendations about how to use our TDF results and the COM-B model to design interventions and behavior change. In sum, the therapists provided support for the implementation of BT for depression in routine primary care. These five therapists advocated blending online sessions with f2f therapy as an innovative care form and saw the possibility of combining the best of both in their daily work.

76The following is the supplementary data related to this article.

77None.

78The current study was funded by the European Commission FP7-Health-2013-Innovation-1 program (grant agreement number: 603098) and was conducted by the German trial partner of the project European Comparative Effectiveness Research on Internetbased Depression Treatment (E-COMPARED). The funder had no role in research idea, study design, data collection, analysis and interpretation, decision to publish, or preparation of the manuscript. We thank the therapists and the second coder Steffen Hartmann (SH) for their contributions. Results from the E-COMPARED project can be found at www.e-compared.eu. We acknowledge support by Deutsche Forschungsgemeinschaft and Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU) within the funding programme Open Access Publishing.

79IT designed the study, developed the interview guide, collected the data and supervised/contributed to the data analysis and interpretation as well as drafted the manuscript. KS executed data transcription, performed data analysis and interpretation with a second independent coder and contributed to the draft of methods and results. DE supervised the primary stage of developing the study design and the further writing of the manuscript draft, MB and HR critically revised the manuscript.