An explanatory sequential mixed methods study was conducted in 2 phases: first collecting and analyzing quantitative data, followed by qualitative data collection and analysis [24]. A mixed methods research study has the advantage of being able to elucidate complex issues [25], and using this design, researchers can combine the strengths of 2 approaches and thereby contribute to interdisciplinary and collaborative research evidence [26]. In our study, 2 methods (quantitative questionnaire and interviews) were organized sequentially [27] and analyzed, in accordance with the rules of each method [28].

Adult patients residing in the canton of Vaud, Switzerland, who had an episode of care (inpatient or outpatient) between December 1, 2021, and December 28, 2023, at Lausanne University Hospital (CHUV) with an active EPR were identified (N=839) and matched 1:1 with adult patients without EPR on age, self-identified gender, and care period (N=839); refer to Figure 1. Whether an EPR was present at admission was automatically verified and reconciled against the institutional EHR (CHUV).

A questionnaire (Multimedia Appendix 1) was sent to them, which they returned by post or electronically. All data were collected and managed using REDCap (Research Electronic Data Capture) hosted at CHUV [29]. Demographic data and socioeconomic profiles were collected. A high level of education was equivalent to a higher or specialized school or university. In addition, the level of health literacy and digital health literacy was assessed using a validated questionnaire (HLS19-Q12-CH [Health Literacy Survey 2019-2021] and HLS19-DIGI-CH [Digital Health Literacy], scales in French [30,31]), comparing the profile of this population to a sample of the Swiss population recently studied in 2019‐2021 (Health Literacy Survey Schweiz 2019‐2021 [31]). Voluntary participants provided their contact details for the qualitative phase (Phase 2) at the end of the questionnaire.

On the statistical level, bivariate analyses were performed using 2-tailed Student t test or ANOVA for continuous data, and the chi-square test for categorical data. Bivariate associations were assessed by Spearman correlation. Statistical significance was considered for a 2-sided test with P<.05. Means are expressed with SD. All statistical analyses were performed using Stata version 18 software (StataCorp LLC).

The qualitative analysis was performed to understand the users’ and nonusers’ experiences and perceptions toward the EPR. Semistructured interviews with EPR users and nonusers made it possible to understand the facilitators and barriers as well as the benefits of the EPR, through thematic analysis. The recruitment was structured as follows: participants from Phase 1 provided their contact details at the end of the questionnaire if they were willing to participate in interviews. A list of 140 potential interviewees who responded positively to the invitation (113 EPR users and 27 nonusers) was generated and stratified according to age, self-identified gender, professional status, comorbidities, and health literacy scores. For the user group, a priority list was established to recruit at least 15 participants, both men and women with diverse profiles. The group of nonusers was all invited, but only 8 participants were willing to schedule an interview. The goal was to recruit 20-25 interviewees with diverse profiles among participants, so that a range of perceptions could be collected.

Participants shared their perceptions, their engagement in that experience, and evaluated their experiences regarding the EPR. An interview guide was developed (Multimedia Appendix 2), and pilot interviews were conducted to assess the suitability of the questions. The duration of interviews ranged from 15 minutes to 1 hour 12 minutes (mean 39, SD 17 minutes). To ensure the transparency and validity of the findings, all interviews were conducted in accordance with the interview guide. The interviews were transcribed “ad verbatim” in French (except 1 interview conducted in English) and data were analyzed using thematic analysis [32] with the help of MAXQDA version 24 (VERBI GmbH).

The 6 steps described by Braun and Clarke [32] were used to transcribe the data (step 1), followed by systematically coding the collected data (step 2). The next phase included a compilation of codes into themes (step 3) that were then reviewed (step 4) and discussed. Three researchers (VS, MR, and CL) worked together on the dataset, discussing themes, and refining the relations between the different themes (step 5) and writing it all up (step 6). An Excel table allowed for transparency of the process, being able to compare themes, subthemes, and citations. It must be noted that the reflexive thematic analysis of Braun and Clarke [33], rather than using a codebook approach or a coding reliability approach based on a positivist paradigm [34], was in line with the interpretative analysis of the data.

While the entire data analysis process was performed in French, for this article, themes, subthemes, and citations are translated into English.

At first, data from phases 1 and 2 were analyzed separately and integrated during the interpretation phase to find explanations of the different uses and perceptions of EPR. The integration of the quantitative and qualitative phases happened at 2 levels: first, based on the quantitative results, the selected participants with varied characteristics ensured a rich representation of responders. This approach allowed us to link qualitative and quantitative data at the individual level, and an in-depth analysis of this subgroup was performed. Second, a follow-up analysis was performed based on the quantitative data from Phase 1 [27] to see whether qualitative data were linkable to results from the quantitative data at the aggregate level.

Interpretation was all done in the original language (French), and translation of the quotes into English was performed only for the purpose of authoring this paper.

The concept of reflexivity was adopted throughout the study to ensure its scientific rigor [35]. Specifically, reflexivity is defined as a thoughtful and conscious process that includes continuous evaluation of subjective responses, interpersonal dynamics, and the research process itself [36]. Discussions among the research group members from different disciplinary backgrounds (medicine, physiotherapy, and sociology) facilitated the triangulation of perspectives.

This mixed methods study was conducted in accordance with the Declaration of Helsinki and received ethical approval from the local ethics committee (Commission cantonale d’éthique de la recherche sur l’être humain, CER-VD, no 2023‐01672). Participation was voluntary and all participants were fully informed about the study’s objectives, procedures, and their right to withdraw at any time without consequences. Written informed consent was obtained from all participants prior to completing the questionnaire (Phase 1). For Phase 2, additional informed consent was obtained from the subset of participants selected for in-depth interviews. No compensation was provided. Participant anonymity was maintained throughout the study. All questionnaire responses and interview data were deidentified and assigned unique identification codes. Confidentiality was ensured through secure data storage with password-protected access limited to the research team members.

This mixed methods study provides valuable insights into the characteristics, experiences, and perceptions of EPR users and nonusers in the canton of Vaud, Switzerland. Our findings reveal that early adopters of EPRs in Switzerland demonstrate distinct demographic and health profiles that both align with and diverge from patterns observed in previous international research.

Our results indicate that early EPR adopters in French-speaking Switzerland are predominantly characterized by higher educational levels, chronic health conditions that require regular monitoring, and extensive health care networks. These individuals typically demonstrate higher levels of health literacy and were already actively engaged in managing their health information prior to the adoption of EPR, often through personal archiving systems. These findings align with previous studies that have identified chronic disease status as a significant predictor of EPR adoption [11,18].

However, contrary to international trends that typically show higher female participation in EPR adoption [19-21], our study found a significant male predominance among Swiss early adopters. This gender disparity may reflect broader patterns of digital technology engagement in Switzerland, where technical and IT fields continue to show marked gender disparities characterized by predominant male representation. This unexpected finding warrants further investigation to understand the sociocultural factors influencing gender-based differences in EPR adoption within the Swiss context.

Our study reveals that early EPR adopters demonstrate higher levels of health literacy compared to both nonusers and the general Swiss population. This finding is consistent with previous research that highlights the relationship between health literacy and the adoption of digital health tools [22,23]. The proactive health information-seeking behaviors observed among early adopters suggest that these individuals perceive themselves as legitimate partners in patient-professional relationships, potentially transforming traditional hierarchical dynamics toward more horizontal communication patterns. Similar transformations have been observed in pediatric settings, where parents of chronically ill children demonstrate comparable engagement patterns [39].

Despite recognizing potential benefits, both users and nonusers identified significant barriers to the optimal implementation of EPR. These include technical challenges (such as security procedures and interface design), limited professional participation, and fragmented implementation across the health care system. The critical role of health care professionals as facilitators or barriers to patient EPR adoption emerged as a consistent theme, with many participants perceiving negative attitudes toward EPRs among health care providers. This finding aligns with observations from France and Austria, where professional resistance has similarly hindered EPR implementation [40,41].

Our qualitative findings also revealed concerns about illness stigmatization influencing EPR adoption decisions, particularly for sensitive health conditions. Some participants expressed reluctance to include mental health or gynecological information in their EPRs due to perceived stigma. This observation aligns with German research indicating that patients with stigmatized conditions demonstrate greater resistance to EPR adoption [42] and Norwegian findings showing more critical attitudes toward EPRs among patients with mental health conditions compared to those with physical disorders [43].

The relatively low satisfaction rates among EPR users (30.2%) primarily stem from limitations in system design and implementation rather than the concept itself. EPR users expressed disappointment with the current EPR tool’s functionality, citing a lack of intuitive design, poor document classification, and limited portability. These findings align with global experiences in EPR implementation, where user experience significantly impacts adoption rates and perceived effectiveness [39]. This misalignment of expectations represents a significant challenge for successful implementation.

Our study highlights the importance of considering sociopolitical contexts in EPR implementation. The Swiss opt-in approach to EPR registration, which makes adoption voluntary for citizens but mandatory for newly established health care providers since 2022, has created an implementation gap. This policy approach contrasts with opt-out strategies implemented in countries such as Austria, which have demonstrated different adoption patterns [40]. The influence of professional associations, particularly the perceived resistance from medical societies, emerged as a significant barrier to widespread adoption.

Additionally, our findings suggest that individual factors such as polymorbidity, emotional state, and social support networks influence EPR adoption decisions, consistent with Austrian research demonstrating correlations between these factors and EPR adherence [44].

Our findings have significant implications for the implementation of EPR in Switzerland and similar health care contexts. First, the distinct profile of early adopters suggests that current implementation approaches may be inadvertently creating digital health disparities, with higher-educated, male, chronically ill patients benefiting disproportionately. Targeted strategies to engage underrepresented populations, particularly those with lower health literacy and fewer digital skills, are essential for equitable implementation.

Second, the critical role of health care professionals in facilitating or hindering EPR adoption highlights the need for comprehensive professional engagement strategies. Future implementation efforts to foster greater support for EPR adoption should address professional concerns regarding workflow integration, information accuracy, and time constraints.

Third, the technical and design limitations identified by users suggest the need for user-centered redesign processes that incorporate feedback from diverse stakeholders. Improvements in interface design, document classification, and mobile accessibility could significantly enhance user satisfaction and adoption rates. Participants suggested a wide array of suggestions for improvement (Table 4), useful for policymakers and organizations responsible for implementation. Easier access to the platform (mobile app) and better organization and standardization of documents are needed to respond to the clients’ needs.

Future research should explore the gender disparity observed in our study, investigate effective strategies for engaging underrepresented populations, and evaluate the impact of different policy approaches (opt-in vs opt-out) on EPR adoption patterns in the Swiss context.

Several limitations should be considered when interpreting our findings. The significant gender imbalance in our sample, with male participants substantially overrepresented among both EPR users and study respondents, introduces potential systematic bias. Similarly, the low response rate among nonusers (8.2% compared to 30.4% among users) may have resulted in selection bias, potentially overrepresenting individuals with stronger opinions about EPRs or higher digital literacy. Furthermore, it would have been very interesting to compare respondents to nonrespondents to better understand the population of nonrespondents. However, due to restrictions by the ethics committee and data availability (eg, socioeconomic factors, comorbidities, and literacy), we could not explore these dimensions.

Furthermore, investigating health literacy and digital health literacy using standardized questionnaires (HLS19-Q12 and HLS19-DIGI-CH) can be criticized. A recent scoping review [45] identified 33 studies using digital health literacy tools; however, with several limitations, such as “narrow scope, limited adaptability to the rapidly evolving digital landscape, and lack of objective testing” (p. 11). The choice of the 2 instruments used in this study was based on the possibility to compare them to results from the national survey [31]. This choice aligns well with Thorup et al [45] recommendations to (1) combine health literacy and digital health literacy (multidimensional tool), (2) to use a validated tool, and (3) to use tools that are context specific.

Our qualitative analysis, while providing rich contextual insights, is inherently subjective and may reflect researchers’ interpretations. Despite efforts to ensure analytical rigor through team-based analysis and reflexive practices, alternative interpretations of the data remain possible.

Finally, our study focused on patients with care episodes at a university hospital, which may limit its generalizability to community-based settings or rural populations. The experiences and perceptions of individuals without recent hospital encounters may differ substantially from those captured in our sample.

This study provides valuable insights into the characteristics, experiences, and perceptions of early EPR adopters in the canton of Vaud, Switzerland. Our findings reveal that these early adopters constitute a distinct population characterized by higher education levels, chronic health conditions, extensive health care networks, and higher health literacy. Despite moderate satisfaction with the current EPR implementation, most users recommend the system to others, suggesting belief in its potential value despite existing limitations.

The observed demographic patterns raise important questions about EPR accessibility and potential digital health disparities. The significant male predominance among Swiss early adopters contrasts with international trends and warrants further investigation. Additionally, the higher health literacy levels among adopters suggest that current implementation approaches may inadvertently favor already advantaged populations.

Our mixed methods approach revealed that moderate satisfaction among users primarily stems from limited system uptake across the health care ecosystem rather than inherent opposition to the concept. The optimal effectiveness of EPRs appears contingent upon systemic integration throughout the ambulatory care network, suggesting that extending implementation beyond university hospitals is essential for achieving the intended goals of care continuity and coordination.

As Switzerland continues its digital health transformation, addressing the technical, professional, and policy barriers identified in this study will be crucial for realizing the full potential of EPRs to enhance health care quality, coordination, and patient engagement.