Participants were practicing GPs (specialists in family or internal medicine) working in their own practices or as an employee in a practice, as well as clinical residents. The link to the online survey was distributed nationwide via email to German research practice networks and professional associations, as well as presented at public events for GPs in Germany. Subsequently, the research practice networks forwarded the invitation to their affiliated GPs, resembling a snowball sampling approach. Due to this strategy, it is not possible to determine the exact number of GPs who received the invitation to participate. The data were collected cross-sectionally through an anonymous online survey from January to October 2024.

The questionnaire was developed by the study team according to the Consolidated Framework for Implementation Research model, which captures barriers and enablers in the implementation of interventions [31]. We conducted a pretest of the questionnaire with 6 GPs who verified the relevance and completeness of the items. Based on the feedback of GPs, the wording of 1 question was revised to improve clarity. The overall structure and methodology of the questionnaire were retained after pretest.

We collected sociodemographic information (ie, gender, age, and duration of practicing as GP) and characteristics of GP practices (ie, type of practice [single vs group practice], location [urban-rural], number of treated patients per quarter, and extent of digitization in practice). To determine the extent of digitization in practice, GPs stated on a 5-point scale whether their communication with colleagues in outpatient care and patients is “almost completely digitized” (0) or “almost completely paper-based” (4). For regression analyses, we recoded this variable into 3 categories: “digitized” (0‐1), “half digitized and half paper-based” (2), and “paper-based” (3-4).

Participants were asked how many patients they had prescribed a DHA (all indications) or a DHA-MD in the last quarter. Since a large proportion of participants reported having prescribed no DHA-MD at all, we recoded this variable into a binary measure (prescription=yes or no) to facilitate analysis using logistic regression. A prescription was considered present if at least 1 patient had received a DHA-MD during the last 3 months. The perceived need for DHA-MD for patients was estimated on an 11-point scale, ranging from “very low” (0) to “very high” (10).

Quantitative data were analyzed using SPSS (version 30; IBM Corp). Descriptive statistics were presented as mean values with standard deviations for metric-scaled variables and as percentages and frequencies for nonmetric-scaled variables in order to describe the study sample.

The following variables were the predictors of the regression analyses: gender (male or female), age (years), duration of GP practice (years), type of practice (single practice or joint practice), practice location (city or rural area), and degree of digitization in practice (paper-based, or half digitized and half paper-based, or digitized). Correlations between predictor variables were examined to rule out multicollinearity. In cases of high correlation (r>0.80) [32], a decision was made to exclude one of the variables from the analysis based on theoretical considerations and relevance to the research question. Dependent variables were analyzed via multivariate analyses using logistic regression modeling for prescribing behavior and linear regression modeling for perceived need of DHA-MD. Missing values were not imputed. Participants with missing data were excluded from the respective analyses. The Gauss-Markov assumptions were tested as prerequisites for the multiple linear regression of perceived need of DHA-MD. Model fit was assessed using Nagelkerke’s R² for the logistic regression model and adjusted R² for the multiple linear regression model. The significance level was set at α≤.05.

Invitations for the telephone interviews were distributed via the German research practice network “SaxoForN” [33], which subsequently forwarded the invitation to their affiliated GPs (snowball sampling). Medical assistants (MAs) were also invited to participate in qualitative interviews, as they are closely involved in administrative processes of GP practices and can provide organizational and time relief for GPs in the German health care system [34]. Previous research has shown that patients have a high level of trust in MAs [35], who often serve as the first point of contact for questions or issues that may not be raised during the GP consultation—possibly also with regard to the prescription and use of DHA. Moreover, as certain tasks in primary care are delegated from GPs to MAs [35], we aimed to explore which concrete tasks they take over in relation to DHA—such as assisting in administrative procedures, supporting patient onboarding, or addressing patient inquiries during the usage phase. Including MAs in the qualitative study enabled us to explore their role in the implementation process and capture patient-related challenges they identified, which may extend beyond the GPs” perspective. Only MAs who worked in GP practices with DHA prescriptions were interviewed to ensure whether they were already familiar with the concept of DHA. Interviews were conducted with all interested GPs and MAs who had registered to participate. Telephone interviews took place between July and October 2024 and were conducted by 2 researchers (SaS and SSch) of the study team. Both interviewers were female and health scientists. They had no prior relationship with the participants.

Semistructured interview guides for GPs and MAs were developed according to the Consolidated Framework for Implementation Research framework [31]. The interview guides were pretested by 2 GPs in December 2023. Feedback revealed that no changes were deemed necessary in terms of methodology, structure, or questions.

The interviews with GPs started with an opening question, which transitioned to the following 3 key topics: experiences with DHA (overall; for DHA-MD), attitudes toward DHA-MD, and implementation factors and conditions for the use of DHA-MD in outpatient care. MAs were asked about their tasks related to DHA in GP practice, frequently asked questions of patients related to DHA, and feedback of patients to DHA use.

If necessary, the interviewers asked further questions to go more in-depth on the information the participants provided. Despite the interview guide, participants were encouraged to talk freely without too much interruption from the interviewers. Theoretical saturation was assessed iteratively during data collection by the interviewers and was deemed achieved when interviews no longer yielded novel insights and a sufficient heterogeneity of perspectives had been captured. To ensure a transparent and consensual process, regular team meetings were held to discuss emerging themes and determine the point of saturation. The telephone interviews were audio-recorded and transcribed verbatim. To ensure data protection, all personal data were pseudonymized in the transcription process.

The interviews with GPs and MAs were analyzed using qualitative content analysis according to Mayring [36]. The coding schemes for the interviews were developed using a deductive-inductive approach, based on the previously developed interview guides. Two researchers (SSch and SaS) analyzed the interviews independently with use of the software MAXQDA (version 2020; VERBI Software). The results were subsequently cross-compared, whereby disagreement was discussed until consensus was reached. If necessary, a third senior researcher was consulted.

A total of 149 participants completed the questionnaire, of whom 47.7% (71/149) were male and 52.3% (78/149) were female. As shown in Table 1, the mean age of the respondents was 50.7 (SD 10.5) years. The mean work experience of GP was 15.2 (SD 9.9) years. Most respondents were practice owners (111/149, 74.5%) and located in urban areas (122/148, 82.4%) with various community sizes (Table 1). About half of the respondents were active in single practices without physician colleagues (74/145, 51.1%), while the other half (71/145, 48.9%) worked jointly with at least 1 colleague.

Of the participating GPs, 65.7% (90/137) prescribed at least 1 DHA in the last quarter. The median prescription frequency for DHA (for all indications) per quarter was 2 (IQR 0‐5) and for DHA-MD, it was 1 (IQR 0‐2). Nearly half of the respondents (68/137, 49.6%) did not prescribe any DHA-MD in the last quarter. The median estimated need was 3 (IQR 1‐5).

The multivariate logistic regression included the predictors gender, age, duration of GP practice, type of practice, practice location, and degree of digitization in practice. Correlations between predictor variables indicated a high correlation between age and duration of GP practice (r=0.87; P<.001), justifying the decision to exclude age as a predictor in the regression model. All other correlations were low (r<.25). Working in a half digitized and half paper-based practice (odds ratio 5.133, 95% CI 1.695‐15.542) as well as working in an almost completely digitized practice (odds ratio 3.006, 95% CI 1.296‐6.969) positively predicted prescribing behavior (Table 2). Hosmer-Lemeshow test indicated a good model fit (χ²₈=12.43; P>.05) for the logistic regression model. The model explained 13.9% of the variance (Nagelkerke’s R²=0.139).

The multivariate linear regression analysis included the same predictors as in the logistic regression analysis. The Breusch-Pagan test indicated no heteroscedasticity (P=.45). The duration of GPs’ medical practice (b=−0.057; P=.01) negatively predicted the perceived need, while working in a group practice (b=0.980; P=.02) positively predicted the perceived need for DHA-MD (Table 3). The model explained 8.2% of the variance (corrected R²=0.082; P=.009).

All interviewed GPs stated that they had already prescribed DHA for different indications in practice, including DHA-MD addressing mental disorders. Reasons for prescribing DHA-MD to patients were positive experiences with individual DHA-MD, patient request, the ability for GPs to view the contents before prescribing (GP test access), and the opportunity to offer patients a treatment alternative to existing therapy options.

GPs’ experiences with health insurance companies, as well as with the activation process of DHA-MD, were mixed. Some GPs mentioned occasional problems with health insurance companies, including long waiting times for prescription processing, prescriptions being completely rejected by individual health insurance companies, and technical problems when redeeming them. Others stated that the processes with health insurance companies and the activation process of DHA-MD were straightforward, and patients usually gained access to DHA-MD within a few days (Multimedia Appendix 1).

Feedback from GPs regarding their experiences with the utilization of DHA-MD by patients was heterogeneous. Some GPs reported positive user experiences of patients. Others, however, felt that only a certain group of patients benefited from DHA and used it as intended —namely, those who were particularly motivated and willing to actively engage in the management of their mental health condition. In a few cases, GPs reported that they issued prescriptions, but these were not redeemed at all by patients. GPs stated that patients carefully consider whether treatment with DHA-MD fits their personal context and preferences and whether the content of the DHA-MD aligns with their needs. In addition, GPs reported a lack of integration in the treatment with the DHA-MD, making it difficult for them to monitor patient adherence and the treatment progress.

Regarding the assessment of the effectiveness of DHA-MD, 4 different groups emerged in the analysis. The first group of GPs, who considered DHA-MD to be effective and without side effects, emphasized that DHA-MD would have stabilizing effects on patients by providing validated knowledge about their mental disorder and is especially helpful in bridging the waiting time for psychotherapy appointments. GPs further stated that DHA-MD could reduce physician-patient contacts and, in some mild cases, make psychotherapeutic treatment no longer necessary. The second group of GPs stated that DHA-MD effectiveness depends on the patient characteristics. Patients must be motivated and should use the DHA-MD as prescribed by the manufacturers to achieve sufficient effectiveness. The third group of GPs declared that DHA-MD effectiveness depends on the indication. According to the GPs, DHA-MD would be effective for conditions where cognitive-behavioral therapy is effective, such as anxiety disorders, but less so for borderline personality disorders. In addition, GPs mentioned that providing proof of effectiveness might be more difficult for certain mental health conditions because it cannot be measured as objectively as in other conditions (eg, weight loss in obesity). The fourth group of GPs expressed that evidence-based statements about the effectiveness of DHA-MD cannot be made due to insufficient evidence. They highlighted the lack of long-term studies and large-scale cohort studies, including subgroup analyses, to assess effectiveness in different patient groups.

With regard to the importance of DHA-MD in primary care, GPs mentioned societal benefits, meaning that the treatment with DHA-MD could reduce the use of antidepressants, minimize sick leave, and ease the burden on GP practices by reducing patient-physician contacts.

Regarding the question of how GPs assessed the potential of DHA-MD in the collaboration between GPs and psychotherapists, 2 different groups emerged. The first group of GPs emphasized that there has been little exchange between GPs and psychotherapists so far, and they believed that this situation would not change even with the development and integration of these innovative digital technologies into patient care. A second group of GPs could imagine that both groups of health care providers would be included in the DHA-MD, allowing them to simultaneously track the treatment process and promote interdisciplinary collaboration.

GPs’ assessment of the perceived need for DHA-MD in primary care was heterogeneous. On the one hand, GPs stated that there is a need for DHA-MD, justified particularly due to an increase in mental disorders, especially among younger adults, and as a consequence of the COVID-19 pandemic. In addition, they reported that DHA-MD would be helpful if they addressed a problem in general practice—for example, by providing treatment for patients in need of psychotherapy who face long waiting times, thereby alleviating pressure on GP practices. However, GPs also highlighted that developing new DHA-MD for indications already covered by existing products is unnecessary, as the market is becoming increasingly opaque, and they lack the time in their daily practice to inform themselves comprehensively about new products. Instead, they advocated in the interviews for a reevaluation and regular improvement of the quality of existing DHA-MD. On the other hand, GPs stated that the evidence for practical use of DHA-MD is still insufficient and that these digital applications would not be necessary if there were enough psychotherapy places and enough medical capacity that the conversations with patients could be partly conducted by GPs themselves.

Overall, MAs’ experiences with DHA were heterogeneous (Multimedia Appendix 2). All interviewed MAs stated that DHA had already been prescribed in the GP practices where they worked, including DHA for musculoskeletal disorders, mental disorders, and metabolic disorders. According to MAs, their main sources of information about DHA were the internet, manufacturer advertising, specialist journals, and test access. However, all MAs reported that they lacked sufficient knowledge about the contents and functionalities of DHA and expressed a need for increased public relations work and training in order to better assess the functionality of DHA, as well as the suitability for individual patients. MAs also reported challenges in selecting patients suitable for DHA-MD use because patients often do not talk openly about their mental disorder due to feelings of shame.

MAs evaluated the importance of DHA in primary care differently. On the one hand, DHAs were perceived as an innovative care solution, particularly for bridging waiting times for psychotherapy and as a supportive therapy. On the other hand, doubts were expressed about their usefulness, since personal contact to a GP or a psychotherapist was considered essential for mental disorders. With regard to the integration of DHA into GP practice, MAs reported that implementation in rural areas with a predominantly older patient population is difficult because they tend to be less tech-savvy. Most MAs reported not taking on any tasks related to DHA in GP practice. In some cases, there were organizational tasks for MAs in relation to DHA, for instance, preparing flyers for interested patients or reminding GPs to use the billing code in the reimbursement process when a DHA was prescribed. Most MAs reported that they had not been asked any questions by patients in connection with DHA. There were single questions regarding the functionality of DHA, the necessity to use, and if patients specifically requested a DHA.

According to MAs, detailed questions of patients were addressed directly with the GPs during the consultations. Regarding the feedback of patients to DHA, which was shared with MAs, three different groups emerged: (1) positive feedback, (2) negative feedback, and (3) no feedback received. The first group reported positive user experiences. According to MAs, younger patients who were highly motivated to use DHA benefit particularly and receive support in dealing with their own disorder, as well as making sustainable lifestyle changes. The second group reported negative feedback in detail that patients did not get along with the use of DHA. There was a third group of MAs who stated that they did not receive any feedback from patients because this was discussed exclusively during consultations between the GP and the patient.