The study was conducted as part of the regional gynecological specialist training days on breast cancer care (GYN-Fortbildungstage Mammakarzinom) in Marburg, Germany on December 6, 2023. The event encompassed presentations about current developments in breast cancer care, i.e., targeted therapies for breast cancer including antibody-drug conjugates or changes in reimbursable disease management programs for breast cancer patients. Only caregivers which were actively involved in patient care of the regional breast cancer care network of the county of Marburg-Biedenkopf did attend. Recruitment for the study was based on voluntary participation in the survey by the attendees of the mentioned event. An ethics vote was waived by the Research Ethics Committee of Philipps-University Marburg on November 13, 2023 (23-279 ANZ) with reference to the necessity of primary data anonymization, neglecting the option for follow-up survey. Participants who did not want to take part in the survey were able to leave the room beforehand. No one took advantage of this option, all attending participants in the event took part. The requirement for inclusion in the study was the active participation in the regional breast cancer care network and patient care. Exclusion was performed for participants who do not have a medical profession (e.g., administrative employees). The comprehensive survey was initially shared with attendees via a QR-code, with responses gathered directly through Google Forms (Google LLC, Mountain View, United States). Following the survey, participants underwent a 30 min educational presentation about the application of artificial intelligence and blockchain in managing breast cancer. The presentation included the demonstration of previously publicized concepts on combined application of artificial intelligence, i.e., the use of large language models as an adjunct for tumor boards, as well as blockchain technology, i.e., as technological framework for decentralized interoperable data management in care networks (8, 10, 30). After the presentation, two questions concerning the intended use of these technologies were posed once more.

The study encompassed physician and non-physician healthcare professionals involved in breast cancer care within the regional network under investigation. The preliminary set of questions aimed to gather fundamental demographic data, including age, gender, and details concerning their practice in either outpatient or inpatient settings, in addition to their level of professional training.

The initial survey consisted of 22 items (see Supplementary material S1 for original German version), beginning with four questions to capture basic demographic details of the participants as described in 2.2. This was followed by a series of 10 questions evaluating the participants’ awareness and intention-to-use of three technologies (HCP): (1) digital health applications (DHA), (2) artificial intelligence (AI), and (3) blockchain technology (BC). Technology awareness was measured in binominal manner (yes or no) while intention-to-use was assessed using a 7-point Likert scale (1 = strongly disagree; 7 = strongly agree). The survey then addressed the components of the German version of the eHealth Literacy Scale (GR-eHEALS 1–8) (31, 32). Table 1 is provided to summarize the questionnaire and the abbreviations of its items.

The eHealth Literacy Scale (eHEALS), originally introduced by Norman and Skinner in 2006, has since been adapted into various languages and validated across multiple settings (32). The 8-item scale is used for assessing electronic health literacy in research populations, employing a 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree), with total scores spanning from 8 to 40. Higher scores denote greater literacy levels. In our study, we utilized the validated German version of the eHEALS (GR-eHEALS) by Marsall et al. (31). This study marks the first application of GR-eHEALS among health professionals in the field of gynecology, prompting us to examine reliability and validity in this novel context in line with Norman and Skinner. Internal consistency was measured through Cronbach’s alpha. Validity was examined using exploratory factor analysis. We employed the maximum likelihood method, adopting factors with an eigenvalue exceeding 1, in line with the Kaiser criterion. The decision on the number of factors to retain was based on the analysis of a scree plot.

Data analysis was conducted in IBM SPSS Statistics (Version 29.0.2.0 (20); IBM Corporation, Armonk, United States). We computed Spearman’s rho to establish nonparametric correlations for the variables of age, ITU_AI, ITU_BC, and eHEALS_Total, and these were then evaluated for their statistical two-tailed significance. Additionally, we carried out Mann–Whitney Test to compare ITU_AI and ITU_BC scores across binary categories of gender and sector.

Following the educational session, the intention-to-use query was repeated for the investigated technologies (ITU_AI; ITU_BC). Differences in pre-and post-comparison were assessed by conducting a two-tailed Sign test on the two variable pairings (ITU_AI_PRE 🡢 ITU_AI_POST; ITU_BC_PRE 🡢 ITU_BC_POST).

The study population is characterized by a gender distribution of 57.8% female {mean age (SD): 48.26 (±13.96)} and 42.2% male {42.50 (±11.20)} for 45 participants {44.93 (±12.62), minimum 24 to maximum 67 years}. In terms of sectoral engagement, 35.6% are predominantly involved in outpatient care, while a majority of 64.4% operates within an inpatient setting. With respect to professional qualifications, the cohort comprises one-third specialists, accompanied by 24.4% of doctors in residency and 17.8% serving as attending physicians (see Figure 1).

Figure 2 illustrates the reported awareness that participants have with the technologies being examined. DHA (Digitale Gesundheitsanwendungen, DiGAs) were the most recognized, with 68.9% awareness, followed closely by artificial intelligence at 66.7%, while blockchain technology as the least known, with only 24.4% of participants being aware of it. Regarding the specific utilization of these technologies in the healthcare sector or in the context of breast cancer treatment, the awareness rates were 51.1 and 26.7% for AI, and for blockchain technology, the rates were 17.8 and 13.3%, respectively.

The initial intention-to-use, as measured by a 7-point Likert scale, yielded a mean of 5.42 (SD; ±1.82) for artificial intelligence and 4.16 (±2.04) for blockchain technology. Figure 3 depicts the distribution frequencies corresponding to the respective scores and types of technology. Moreover, the analysis did not reveal any statistically significant variances in the intention-to-use when comparing across the binary classifications of gender and sector for either technology.

Spearman’s rank correlation coefficient (ρ) was employed to evaluate the significance of associations among the variables: age, intention-to-use artificial intelligence (ITU_AI), intention-to-use blockchain technology (ITU_BC), and the aggregate scores of the German eHealth Literacy Scale (GR_eHEALS_Total). Consistent with the thresholds defined by Cohen et al., a robust positive correlation was identified between ITU_AI and ITU_BC (ρ = 0.591; p < 0.01) (33). Concomitantly, an intermediate-level negative correlation was discerned between participant age and ITU_AI (ρ = −0.438; p < 0.01), while an intermediate-level positive correlation was found between ITU_AI and GR_eHEALS_Total (ρ = 0.383; p < 0.01) (Table 2).

For the exploratory factor analysis, significant findings from Bartlett’s test of sphericity (χ 2 = 304.000, p < 0.001) supported the factorability of the correlation matrix. The high value of Kaiser-Meyer-Olkin test (0.848) showed adequate sampling. Maximum likelihood method confirmed a single factor, which emerged from an initial eigenvalue of 5.47 and explained 68.37% of the total variance explained for. The factor loadings for this model were between 0.619 and 0.906. The data yielded a mean accumulated score for GR-eHEALS of 33.04 (SD ± 6.61; 95% CI {31.06, 35.03}) with a calculated Cronbach’s alpha amounting to 0.928. Detailed results of the exploratory factor analysis and the corresponding scree plot are presented in Supplementary material S2, S3.

With regard to the post-education survey conducted, 30 respondents repeated the question on intention-to-use AI or BC on a 7-point Likert scale. In this sub-group, the mean intention-to-use AI was 5.37 (±1.81) before the session and 5.83 (±1.64) afterwards, which did not show a statistically significant change. The intended usage of BC did significantly increase from an initial average of 4.30 (±2.04) to 5.90 (±1.67), marking an improvement of +1.6 (p < 0.01) (Figure 4).

Public awareness and understanding of emerging information technologies vary and their application often stays unnoticed (15, 18, 37). Despite evidenced benefits for clinical care, a gap between the digital technologies´ potential and actual implementation persists (24). The dissemination of innovative technologies in healthcare necessitates an examination of how professional training can help to close this gap. A study conducted by the Commission for Digital Medicine of the German Society for Gynecology and Obstetrics underlines a significant incongruity between the perceived benefits of digital medicine and its practical application in gynecological care in Germany (14). Despite 78.4% of gynecologists recognizing the ability of digital medicine to reduce their increasing workload, only 13.5% acknowledge receiving institutional training. The research identifies critical obstacles to the adoption of digital medical technologies, primarily grounded in the scarcity of time and a deficiency in knowledge. This study illustrates that even a short training session of thirty minutes can significantly increase the intention-to-use less-known technologies like blockchain, elevating its intended use to the same level as that of more familiar technologies such as artificial intelligence.

eHealth literacy refers to the capacity of individuals to use emerging information and communications technologies for the enhancement or facilitation of health and healthcare services (38). Previous research has established the eHEALS as a reliable and valid measure of eHealth literacy across different contexts (32). This study represents the first utilization of the German version of eHEALS (GR-eHEALS) to delineate the eHealth literacy of healthcare professionals in gynecological oncology, affirming the tool’s applicability in this context (31). Additionally, the positive correlation between eHealth literacy scores and the intended use of AI reinforces the tool’s validity in assessing competencies relevant to digital medicine. Our findings align with past research, which found a positive correlation of eHEALS with internet usage and reported computer knowledge (38, 39). The study confirms a sufficient degree of eHealth literacy among the investigated gyne-oncological care professionals (31, 32). This suggests that healthcare professionals in breast cancer care possess substantial digital skills to use emerging information technologies.

Despite demonstrating a satisfactory level of e Health literacy, adoption rates of emerging technologies often fall short of expectations (14, 24). In a survey conducted by the German Society of Gynecology and Obstetrics, gynecological professionals were asked about their actual use of DHA (Digitale Gesundheitsanwendung, DiGA) and a mere 10.2% of gynecologists had ever prescribed an app (14). Nearly half of the questioned professionals were not even aware that these applications were available for prescription, despite them being reimbursable in Germany since 2019.

This study suggests that the prescription rate of DHA among physicians has increased and provides further insights on less common emerging information technologies of AI and BC. As such their awareness in healthcare and breast cancer care fall behind the DHA. Despite the growing body of evidence underscoring their clinical applicability, lower levels of awareness among health professionals regarding their use in the management of breast cancer persist.

These differences are not only observable in the context of technology awareness but are particularly pronounced in their delineated intended use for breast cancer care. AI is more widely recognized and has a higher intended use for breast cancer care in comparison to BC. Furthermore, the analysis reveals that younger individuals or those with higher levels of eHealth literacy show a greater inclination towards using AI – an observation that seems to be intuitively logic at first glance. Nevertheless, these observations are not consistently applicable to the less known technology of Blockchain. Professional education of 30 min did boost the intention-to-use both AI and BC, with the impact being particularly noteworthy and significant in the case of BC. As a result, after the educational presentation the intended adoption of BC equalizes with the level of intended adoption of AI in breast cancer care. The incorporation of AI in healthcare helps with diagnostic support, patient interaction and treatment customization. The effectiveness of diagnosis and treatment is usually the focus of scientific analysis. However, for a successful integration of the technology into clinical practice, psychological and social issues such as trust in AI, dependency risks and changes in the doctor-patient relationship must also be considered as suggested by Triberti et al. (40). The research by Strika et al. and Sebri et al. highlights the importance of understanding the attitudes of healthcare professionals and patients towards AI, an aspect that is often overlooked in research (27, 29). They argue that future studies should examine the broader impact of AI on the social and organizational aspects of healthcare, not just its effectiveness in diagnosis and treatment.

The research utilized a single-center approach, limiting the wider applicability and relevance of its findings beyond the context of the German healthcare system and its peculiarities in regulatory and socioeconomic dimensions. Unlike Germany, which has established the DiGA process to manage the prerequisites for DHAs´ reimbursability, other European nations do not offer standardized frameworks. As such, the equivalent term of DiGA for a DHA in Germany, needs to be interpreted in a different light compared to other European partner countries.

In a national context, the study’s focus on a single region could be expanded to include multiple and diverse regions across Germany to enhance the national relevance of the findings. Nevertheless, this study, conducted during regional training on breast cancer care in Hesse, ensures a representative sample by interviewing a balanced mix of participants. The study’s demographic balance across gender and professional training, coupled with the high centralization of care and the socioeconomically balanced, predominantly rural area of Marburg-Biedenkopf county, mirrors German standards well enough to provide a solid foundation for this kind of health services research. Previous epidemiological studies of the care network have proven to capture German standards to a sufficient degree (8, 41).

Nevertheless, to improve the external validity and broader applicability of the results, future research should adopt a national multi-network and international approach for comparison. This would enable the study’s findings to be more transferable and applicable to various health systems and care settings. Furthermore, the present study explored the awareness and intention-to-use on conceptional applications of AI and BC, rather than a singular digital application or therapeutic intervention, i.e., an mHealth application. Future research on the adoption of digital health tools should incorporate a study design with follow-up survey to provide more insight into the long-term impact of the educational intervention on physicians and patients perceptions and understanding the factors that hinder or promote adherence to prevent discontinuation of their use.