A qualitative interview study was conducted to gain new insights into the perspectives of psychological psychotherapists regarding their needs, attitudes, and wishes for information about DiGA for mental health purposes in their daily practice. As evidence remains sparse, we chose semi-structured interviews, which facilitate an in-depth understanding of underexplored topics (36, 37). This applies to the needs assessment and wishes regarding information strategies on DiGA. To ensure transparent and complete reporting, we adhered to the Consolidated Criteria for Reporting Qualitative Research (COREQ, see Supplementary Table S1) (38). The Institutional Review Board of the Medical Faculty of the Heinrich Heine University (HHU) Düsseldorf provided ethical approval of the interview study (number 2023-2338) on February 7, 2023. We preregistered the formative substudy of the research project at the Open Science Framework website on June 10, 2024 (https://osf.io/uj8gt).

Individuals were eligible to participate if they were registered as licensed psychological psychotherapists in Germany with the legal option to prescribe DiGA for mental disorders at the expense of statutory health insurance companies (i.e., authorization from the Health Insurance Company's Association). Another inclusion criterion was fluency in German, the language for the interviews. Participants were recruited via social media (e.g., Instagram, LinkedIn) and German professional associations (e.g., the Federal Association of Psychotherapists Chamber, the Federal Association of Contract Psychotherapists, the German Psychotherapists Network). The recruitment material consisted of postings, flyers, emails, and a short video. It contained brief information about the study and its inclusion criteria, along with a link to sign up. Psychotherapists who expressed interest in participating were provided with detailed study information, the informed consent form, and a brief background questionnaire with 13 items on sociodemographic details as well as on DiGA experience (see Supplementary Table S2). A convenience sampling strategy was applied. Participants received a compensation of 100 euros. Of 39 interested therapists, 10 were excluded, not meeting the inclusion criteria. After reaching a point at which it could no longer be assumed that further interviews would provide new insights (i.e., theoretical saturation), 16 licensed interested psychotherapists meeting the inclusion criteria were declined (see Supplementary Figure S1).

Interviews were conducted online using the videoconferencing software Webex (Cisco Systems, CA, USA), with access provided through the institution (HHU) for university staff to facilitate recruitment across Germany. Participants were advised not to enter personal data into Webex and could decline using the webcam. The video was not recorded; instead, audio files were recorded using a voice recorder without an internet connection. Participants could also choose to be interviewed by telephone (via the University Hospital Düsseldorf's conventional telephone network); however, this option was not utilized. Pilot interviews with three female psychologists were conducted to test the technical implementation and the content-related process (39–41). The interview guide (see Supplementary Table S3) was developed by JAH, grounded on empirical research (42), and discussed within the study team. It was slightly adjusted after the pretests and the first interviews by KP and CW. Online interviews were conducted from July 5 to August 16, 2024, by KP (n = 10) and CW (n = 3), held in German. Both interviewers, female doctoral students, were trained in qualitative interview methods beforehand. The interviewers had no contact with the participants before recruitment. They introduced themselves as a psychologist (KP) and a medical graduate (CW) before the interview started, briefly explaining their personal interest in DMHIs. Only the participant and the interviewer were present during the online interview. Interviews were audio-recorded, pseudonymized, and transcribed verbatim using artificial intelligence with manual control compliant with general data protection regulations. The transcriptions were revised and checked for accuracy by the respective interviewers before being analyzed. Participants were not asked to provide feedback on the transcripts. However, until anonymization, they had the opportunity to review their transcripts (as documented in the study information). After analyzing the transcriptions, all audio records were deleted. Likewise, the interviewers' field notes were shredded. After writing the results, the pseudonymization key was deleted, and transcriptions were anonymized.

Interviews were conducted until theoretical saturation (i.e., the point at which no further information gain is to be expected) (43, 44), which was first discussed after nine interviews and considered fully achieved after thirteen interviews.

Building on Kuckartz and Rädiker (45), a content-structured qualitative content analysis with an inductive-deductive approach was conducted. The transcripts were coded and analyzed in several steps. Based on the semi-structured interview guide, KP and CW developed the main categories, which were discussed with the principal investigator (JAH) of the mixed-methods study, an expert in the field of the determinants of the acceptance and use of DMHIs. They then coded three interviews independently to develop subcategories inductively, which the senior researchers JAH and AL, with thematic and methodological expertise, reviewed. After another round of independent coding of two additional interviews, the doctoral students KP and CW iteratively adjusted the coding system. JAH and FS subsequently reviewed the coding system, suggested refinements, and then approved this final coding system, after which KP analyzed the interviews. The final coding system was further discussed with other project team members, including experienced researchers on provider research (NKS, MV, MZ) and qualitative research (LG, JB). The research team members involved in the process come from diverse training and academic backgrounds, including psychology (KP, FS, CP, JAH), medicine (CW, PA), health economics (NKS, MV, MZ), health services research (AL), public health (LG, JB, AL, CP), epidemiology (AL, CP), occupational medicine in health sciences (AL, JAH), medical sociology (JB, CP), and literary translation (LG). This ensures intersubjective transparency, replicability, and discriminatory power of categories and levels. The coding scheme contained three levels, with deductive and inductive categories (Supplementary Table S4 shows, for clarity, two of the three levels with anchor examples).

In total, n = 13 online interviews with psychological psychotherapists were conducted and analyzed (i.e., no dropouts), lasting 32.53 min on average (range: 25.19–43.10 min). As shown in Table 1, the mean age of the participants was 38 years (M = 38.08, SD = 5.06), and 69.2% (n = 9) identified as women.

Self-reported awareness. In the pre-interview questionnaire, all participants reported having heard of DiGA for mental health purposes or DMHIs. Supplementary Table S5 provides detailed information regarding awareness of DiGA for mental health or DMHIs.

Association with the term. Participants linked DMHIs or DiGA primarily to apps for smartphones, digital-supported therapy, and named some DiGA. They associated techniques across various medical treatment areas, including respiratory issues, mental health, and smoking cessation. From their perspective, it reflects the progressive integration of new media.

Current state of knowledge. Participants provided detailed insights into their understanding of DiGA. Most described their knowledge as fairly moderate, often citing a perceived need to be adequately informed to prescribe and guide the responsible use of DiGA. Some participants indicated they were well-informed, typically through part-time research activities or involvement in a collaborative developmental process. Conversely, some participants felt poorly informed, indicating they only knew the mere existence of DiGA or had never worked with them.

Previous information channels. According to how they obtained their current knowledge, some participants reported having received education on DMHIs or DiGA during their studies or in their psychotherapeutic training. Other participants noted that their university programs included relevant content, and only one mentioned having attended specific topic blocks focused on online therapy. Some participants engaged in their own inquiry, driven by intrinsic motivation, personal interest, or a desire to learn more about DiGA:

Participants mentioned having already experienced collegial exchanges, such as through quality circles or intervision groups [collegial supervision, i.e., case discussions among colleagues], but the extent of information gained varied among them. Information from professional associations or the Chamber of Psychotherapists (Association of Statutory Health Insurance Physicians) was rarely cited as a source of information. Some participants noted receiving information through advertisements and the providers' websites, which offered them comprehensive details about DiGA.

Previous prescription practice (behavior). In the pre-interview questionnaire, over 60% of participants indicated having prescribed a DiGA at least once, and 77% had given advice on DiGA or DMHIs. Of those participants who had already been prescribed DiGA, one admitted to having done it only selectively. Few mentioned having recommended DiGA several times and reported that it was not really accepted by their patients.

Effort and practicability of the prescription process. Participants shared varying experiences regarding the effort required to recommend or prescribe DiGA. They reported needing between 2 and 30 min to complete the prescription process, partly including patient education. Some mentioned spending additional time beforehand to familiarize themselves with the process, with one interviewee indicating that this took 2–4 h. One participant noted referring patients to a general practitioner because the relevant feature had not yet been integrated into their practice software. It was stated that the amount of effort required depended on everyone's prior knowledge of the process:

Feedback from patients and observed usage behavior of patients. The responses regarding feedback or prior use by patients varied significantly. Participants reported having little to no feedback; for example, they did not ask about it by default, prescribed it for outsourcing purposes, or had not seen the patient since then. Others received positive feedback, noting that the presentation was engaging, well-structured, and effectively conveyed information. Additionally, participants mentioned receiving negative feedback from patients, indicating that the setup was complicated, personalization was insufficient, demands were too high, or patients did not use the DiGA after downloading it:

One participant suggested an option to familiarize patients with DiGA:

We specifically asked participants about areas in which they believed they could improve their understanding of DiGA, and factors motivating them to seek more information or boost the prescription of DiGA. Additionally, we inquired about perceived benefits and barriers related to prescribing DiGA.

Need for information improvement. Participants suggested that enhancing the dissemination of information could focus on technical aspects. They identified specific gaps in their ability to prescribe DiGA, noting that a comprehensive guideline would be beneficial. In addition, they expressed uncertainty about when to issue prescriptions and how to integrate this process into the routine practice effectively.

Incentives to look for more information or prescribe DiGA. Participants acknowledged that numerous obligations and limited time were key barriers to prescribing DiGA. They noted that one advantage of engaging with DiGA would be improved compensation, such as financial remuneration, Continuing Medical Education (CME) points, and free training. They also mentioned that seeing patients succeed through a DiGA would be an additional incentive.

Perceived benefits of DiGA. According to the participants, the advantages of DiGA include bridging waiting periods and serving as a supplement or partial substitute for face-to-face therapy. This can be particularly beneficial for improving initial motivation, especially in rural areas, and for outsourcing psychoeducation, knowledge transfer, or exposure therapy. From their perspective, DiGA also aids in aftercare or relapse prevention. One participant noted that DiGA could facilitate patient access, especially for those who may feel hesitant or face difficulties seeking therapy. Others mentioned that DiGA could serve as support after basic care, highlighting the usefulness of special journal functions.

Perceived barriers of DiGA. Participants highlighted several barriers to recommending or prescribing DiGA. It was noted that, despite having a positive mindset, the lack of established routines seemed to hinder their use. Few participants expressed skepticism about using DiGA as a substitute or temporary alternative for face-to-face therapy, arguing that it might lead to an underestimation of psychotherapy's effectiveness, prevent patients from beginning person therapy, and may not adequately address complex disorders. Another concern raised was the risk of overwhelming patients, particularly due to unsuitable applications or insufficient personal support. Some participants worried that transitioning therapy to the digital realm could lead to the elimination of their current practices due to competition or lower costs. There were also concerns regarding managing upcoming patient's crises effectively:

Participants also expressed concerns about potentially facing recourse claims¹ for their DiGA prescriptions.

To better understand their information needs and wishes, participants were asked about the essential content required for making informed decisions regarding the prescription of DiGA for mental health purposes.

Evidence and efficacy. All participants preferred obtaining information about the evidence and efficacy, with varying levels of detail. This information should include an updated meta-analysis addressing evidence and be freely and easily accessible in the register. Some participants felt confident that the admission regulations of DiGA had been thoroughly tested.

Data protection. Participants expressed that information on data privacy was a priority for them. They wanted to be informed about server locations and types of data processing. However, other participants indicated that they trusted the approval regulations of DiGA.

Cost information. Participants were interested in information about costs for various reasons. However, they emphasized the need for transparency regarding this issue. They wanted to communicate the value of the therapy to patients, assess its potential impact on the healthcare system, compare different DiGA, or address health and professional policy reasons:

User-friendliness. Participants highlighted the importance of strong usability, including availability across various platforms, language options, geriatric psychiatric aspects, accessibility settings for individuals with (learning) disabilities, and user numbers to be able to evaluate adherence.

App structure and design. Understanding the specific structures was essential for making informed prescribing decisions for participants. They wanted to know about various key elements, such as the availability of chat features, components for activity development, cognitive restructuring, duration of use, length of the program, individualization options (e.g., adjustment of the speed), own possible contact person (e.g., psychologist), push notifications, and reminders. The information about therapeutic principles (e.g., manual-based, in accordance with guidelines) seemed crucial to ensure alignment with the therapeutic approach or determine how the tool could be utilized and integrated into their therapy practice:

Participants stated it would be helpful for them to receive examples of the apps, such as screenshots of certain features.

Other prescription-relevant factors. Participants mentioned additional relevant factors to make responsible recommendations or prescriptions, such as indications, contraindications, or specific socio-demographic information of patients. It was proposed that information could be similar to the yellow list of the pharmaceutical industry. Further, it would be helpful for some participants if this information were presented in a clear, tabular format that compares different applications, highlighting their advantages and disadvantages.

Prescription modalities. Participants suggested adaptations and adding information to the existing prescription procedure, such as implementation in the clinic or practice software and prescription forms:

Updates on certain DiGA. Participants suggested that receiving information about the latest updates could lead to a reassessment of a DiGA that was previously deemed uninteresting.

Developer/provider background. To evaluate the professional background effectively, participants found it important to obtain detailed information about the provider or developer team of certain DiGA.

Information sources. Participants identified the German Chamber of Psychotherapists, the Association of Statutory Health Insurance Physicians, or the various professional associations as reliable sources of information. From their perspective, these organizations serve as the primary contact points, providing all essential information relevant to their work and therapy. Additionally, respondents emphasized the importance of collegial exchange, which includes insights from experienced colleagues, professionals, research, and established guidelines. Furthermore, participants expressed interest in receiving information from legislators (e.g., the German Federal Ministry of Health, BfArM, DiGA directory), as these entities are responsible for quality assurance. Some participants preferred obtaining information from training or education institutes, seeking neutral perspectives, and receiving CME points. Other participants found testimonials from patients credible, particularly if they came from their patients or relatives. However, participants viewed anonymous evaluations skeptically, as they could be fabricated for promotional purposes. Manufacturers' websites were generally regarded as good sources for insider information, while aggressive advertising emails were viewed negatively, especially if they became overwhelming. Participants thought that health insurance funds and other payers are not a completely trustworthy source of information, as they might have a vested interest:

Other sources mentioned included (public) media, practice management software, and social networks such as Instagram and LinkedIn.

Media channels. Various access forms were mentioned, including virtual options and print materials such as specialist journals, brochures, and manuals. Additionally, digital and postal newsletters or training courses (digital and face-to-face) and online searches in various portals and search engines were mentioned. Participants expressed a desire to get test access for trying out the app to have a better understanding:

Willingness to invest in information acquisition. For an initial search, participants were willing to spend 5–10 min for a quick overview or first exploration of the topic. If the search had a specific purpose, this time could extend up to 2–3 h, depending on prior knowledge and experience. For further training, participants were ready to invest from 1 h up to 2 days, depending on their actual workload and personal circumstances, if they saw added value for their work, and the content was rich. Generally, participants were not inclined to pay for further training on DiGA, given that the system was, among other things, designed to alleviate their workload. However, depending on the training's depth, duration, and delivery mode (e.g., in-person, virtual, live, or recorded), they might be willing to invest between 50 and 350 euros.

Attitudes toward artificial intelligence (AI). We also asked participants for their opinion on receiving information about the latest developments concerning AI and specific wishes on how to be informed most effectively. Participants emphasized the importance of AI in their field, stating that it might ease everyday practice. Concerns were raised regarding, e.g., data protection or patient safety. The increasing relevance also seems to arise from patients addressing this issue in therapy:

Participants preferred digital information, free of charge (e.g., through email), that sheds light on various aspects of AI or multi-sectional articles to show the latest developments, supported, for example, by the BfArM, the Chamber of Psychotherapists, or the German Federal Ministry of Health.

One strength of this study is that we managed to reach a relatively diverse group of participants. Our interviews included psychological psychotherapists with varying levels of work experience, ranging from 4 to 21 years, and differing educational backgrounds. Notably, one psychotherapist expressed no interest in prescribing DiGA. At the same time, our sample distribution, in terms of gender and age, closely aligns with the population of psychotherapists in Germany (72). We are among the first to release specific qualitative study results for this target group in fall 2024, providing current data after 4 years of real-world experience since the implementation of the DiGA registry of the BfARM. Our study also benefits from the diverse backgrounds of our research team, which was mixed in terms of genders, disciplines (e.g., psychology, medicine, economics, public health), age, as well as professional and academic experiences. This feature likely broadens the range of views on the topic during the conceptual development and contextualization of our findings into a broader healthcare perspective.

When interpreting the findings, it is important to consider the following limitations. Selection bias may have occurred due to the convenient recruitment of participants, who were accepted in chronological order. Regarding issues related to self-selection, it should be considered that, especially in the field of DMHIs, participants with overall positive attitudes or interest in the topic appear to be more likely to participate in such studies (73–75). Using purposive sampling—focusing on factors such as experience, education, and location—could have provided a greater diversity of participants. Another relevant limitation might be that two-thirds of our participants had already been prescribed a DiGA, and all were aware of DiGA or DMHIs. It is unclear whether this high level of awareness results from the 4 years of DiGA implementation at the time of the interviews or if it is influenced by recruitment-related bias, with participants being particularly eager to discuss this topic (e.g., via LinkedIn as one recruitment channel, where we also promoted the study through our personal connections and postings).

The predominantly positive attitude towards DiGA that emerged in our analysis may not fully represent the views of psychotherapists who are less digitally engaged (e.g., older therapists, infrequent users of digital tools, or with limited digital health experience) (34). It may also under-represent therapists who do not use a behavioral (CBT) approach and are likely to be skeptical of DiGA, since CBT is the therapeutic modality most used in the currently listed DiGA (76). Therefore, the transferability of our findings to these less digitally engaged or non-CBT therapists should be interpreted with caution. Future studies should sample participants based on a range of experience with DiGA and digital literacy, including a larger group of psychotherapists unaware of these interventions, as this could provide a more comprehensive perspective. To obtain even broader and generalizable results, future studies should also include therapists with varying attitudes regarding DiGA, psychotherapists in training, and therapists employing different treatment approaches, such as analytical, systemic, or psychodynamic therapy. Finally, social desirability may have influenced the responses by the surveyed psychotherapists.

Future studies should focus on determining tailored information that, for example, addresses the barriers faced by psychotherapists to facilitate the possible adoption of DiGA (77). Likewise, further research is needed to evaluate special core competencies, such as integrating DiGA or DMHIs into the workflow, for determining useful training to support psychotherapists' perceptions (78, 79).

Recent studies suggest that experienced psychotherapists are generally more supportive of the potential benefits of DMHIs than non-experienced ones (80). Additionally, they are less likely to agree with the potential drawbacks associated with these interventions. It would be valuable to investigate whether increased experience with DiGA correlates with a more favorable perspective of prescribing them. Future research should examine confounding variables, such as age, gender, and workplace (rural/urban), which accounted for only a small amount of variance.

Future research should have more focus on evaluating psychotherapists' experience prescribing DiGA, as our study did not quantify how often they prescribe DiGA nor specifically ask how they have benefited patients. Our findings indicate that increased knowledge may lead to greater acceptance and adoption of DiGA, suggesting that tailored information is essential. With more knowledge and experience, psychotherapists may be more inclined to consider DiGA as a support or supplementary therapy, as we learn from other disciplines (81–83). Additionally, further research is required to determine the level of experience necessary to facilitate the prescription process for psychological psychotherapists in Germany.

In our sample, two-thirds reported already having prescribed at least one DiGA. However, we cannot be certain that all participants were fully familiar with the prescribing process of DiGA. Previous studies show that psychotherapists in Germany tend to have limited knowledge about the prescription procedure (30, 48). It may be beneficial to investigate the act of prescribing itself as a relatively new situation for the psychological psychotherapists in Germany to better understand potential concerns.

As the experience in our target group increases each year, the scope may not only focus on those who are unaware of DiGA, but also on informed decision-making regarding specific aspects (e.g., contraindications) and on how to stay updated. Recruiting psychotherapists who are reluctant to prescribe DiGA or are generally very skeptical concerning the usefulness of DMHIs would be interesting to complement the current findings in an upcoming study, as this would be out of scope for the current mixed-methods study on information needs and preferences among health professionals.

Particular findings from this formative qualitative research on the information needs and content wishes will be used to generate an experimental design for a DCE approach (84).