This study is an implementation evaluation using qualitative interviews to explore the factors determining implementation success, including the enablers and challenges to the app’s uptake and integration into routine care from both patient and clinician perspectives.

The clinic in this study is a private, independent facility specializing in heart rhythm disorders, located within two major private hospitals in metropolitan Queensland. In the Australian context, private clinics are independently funded through patient payments and private health insurance, offering flexibility and quicker access to specialized care outside of the public health care system. It serves patients from local urban populations, as well as those referred from regional and remote areas across the state, spanning over 1.73 million square kilometers [28]. Telehealth facilities are regularly used for consultations irrespective of patients’ residence, and the clinic currently uses both its website and social media to disseminate health education and promote services. A web-based software tool is already available for clinicians to collect and analyze patient data for monitoring clinical progress.

Introduced in April 2023, the mobile app examined in this study is primarily designed to enhance patient engagement in a 12-week cardiac rehabilitation program following catheter ablation for atrial fibrillation. The app allows users to monitor vital signs, record symptoms, and upload electrocardiogram images but does not provide feedback on entered data. It features medication reminders, video and reading educational materials, and guided relaxation videos. Available exclusively to clinic patients free of charge, the app’s effectiveness in this atrial fibrillation cohort had not been formally evaluated. Identifying and managing risk factors are Class I (evidence or general agreement that a treatment or procedure is beneficial and effective) and Level B (data derived from a single randomized clinical trial or large nonrandomized studies) recommendations in atrial fibrillation clinical practice guidelines [29,30].

Patients scheduled for catheter ablation are introduced to the app by clinicians near the procedure date. Interested patients receive email instructions for downloading and signing up. Nurses then use the clinician portal to view enrolled patients and set up a cardiac rehabilitation care plan with clinical data. Post the procedure, nurses promote app use during follow-up calls at 1, 6, and 12 weeks. The care plan is completed at 12 weeks or when the specialist cardiac electrophysiologist is satisfied, but patients can request early discharge. Clinicians access patient data through their portal, and specialists are anticipated to review patients’ progress before the 12-week consultation. Figure 1 illustrates this process flow.

Participants included eligible patients who had been offered the app after catheter ablation, regardless of whether they used it, and clinical staff at the clinic involved in implementing the app. Patients with cognitive and language-related disabilities that might hinder their ability to express opinions were excluded. Clinic nurses informed eligible patients about the study during planned follow-up care calls (6 weeks after ablation) and sent them a participant information sheet. Interested patients were contacted by the research team via email or text to schedule interviews. Nonresponders were identified after no reply to an initial invitation and a reminder after 3 days. Telephone interviews were conducted with verbal consent, and patient participants received AUD 20 (~US $12) gift voucher upon completion.

All specialist cardiac electrophysiologists and nurses involved were eligible to participate, with no specific exclusion criteria. They were initially contacted via email to introduce the study, followed by scheduling interviews with the research team. Clinician interviews were conducted face-to-face, and written consent was obtained.

SH conducted all interviews using semistructured questionnaires (Multimedia Appendix 1), which were developed based on the CFIR [31] and nonadoption abandonment, scale-up, spread, and sustainability—complexity assessment tool interview guide [32]. All interviews took place between April 30 and July 1, 2024.

Recorded interviews were transcribed verbatim using the transcription function in Word (Microsoft). The transcripts were then thoroughly reviewed to confirm accuracy and gain familiarity with the content. Meaningful segments of text were identified, condensed, abstracted, and coded. A constructionist view was adopted during this process, allowing for the exploration of diverse interpretations of stakeholders’ experiences with the mobile app, considering personal, organizational, and contextual factors influencing perceptions of enablers and barriers [33]. The identified codes were then mapped onto the 31 constructs within the five domains of the CFIR: Innovation, Outer Setting, Inner Setting, Individuals, and Implementation Process (Figure 2 [34]).

Each code was initially assigned to a single construct within a domain. In cases where both analysts concurred that a code could fit into multiple constructs across domains, the analysts discussed and reached a consensus on the most appropriate construct. Following this process, codes were read and reread to understand their potential influence on the construct, as well as on the implementation process. Based on this impact, they were categorized as enablers or barriers to mobile app integration into routine health care services. SH and MJA undertook the iterative analytic process collaboratively.

The results of this analysis are presented as a text for each CFIR domain, supported by quotes extracted from interview transcripts. Each quote is assigned a unique ID to ensure participant confidentiality. These IDs distinguish between patients and clinicians but do not specify whether the clinician is a doctor or a nurse to prevent potential identification due to small numbers. Additionally, the IDs do not reveal the management roles of doctors within the organization to further preserve anonymity.

SH has over 8 years of experience as a medical doctor and public health researcher. MJA is an implementation scientist and health services researcher, with a public health background in government and not-for-profit sectors. The remaining members of the research team—WP, TW, SM, SK, and DB—include specialist cardiac electrophysiologists, health services researchers, and health economists, who also contributed to the interpretation of the findings. While SH and SM bring both insider and outsider perspectives due to their experience in cardiology clinical settings and rehabilitation programs, MJA, SK, and DB provided interpretations from an outsider’s perspective, having no direct experience in cardiology. In contrast, WP and TW, with experience in this clinic, contributed insider perspectives.

The study was conducted in conformance with the principles of the National Statement on Ethical Conduct in Human Research [35]. Ethics and administrative approvals were obtained from the human research ethics committee of the Queensland University of Technology and the clinic administration prior to commencing data collection.

Of the 26 eligible patient participants invited to join the study, 18 participants accepted and were interviewed, yielding a response rate of 69%. The mean age of the participants was 69 (SD 9.20) years and 10 (55%) were male participants. Patients exhibited varied use patterns, ranging from completing the cardiac rehabilitation care plan and regularly entering data to never downloading the app (Table 1).

All 6 clinicians involved in the mobile app implementation were interviewed, comprising 2 nurses and 4 specialist cardiac electrophysiologists. The 4 specialists, in addition to their clinical duties, also serve in high and midlevel management positions within the clinic. This dual role allowed them to offer insights shaped by both clinical practice and managerial oversight.

A collective analysis of the codes revealed a complex interaction between key determinants across all five CFIR domains, impacting two main elements in the implementation process: (1) acceptability and sustained engagement, and (2) the clinic’s implementation readiness (Figure 3).

This study contributes to closing the knowledge gap in understanding the context-specific barriers and facilitators that influence the integration of digital health interventions in clinical practice, particularly in the implementation of mobile health apps for cardiac rehabilitation and lifestyle management. In summary, among the many app-related features, a simplified downloading process, intuitive design, and low complexity were key enablers of its acceptability and use, but manual data entry flaws and interoperability issues due to mobile phone type reduced user engagement. Several key patient characteristics also influenced app uptake. Those with already established healthy lifestyle habits (eg, regular exercise) were less likely to find the app useful, while personality traits such as motivation were also found to play a significant role in the adoption and sustained use of the app. Interestingly, older patients were more likely to engage with the app compared to younger patients. Clinicians expressed varied levels of trust in the app’s effectiveness, and patients’ unwillingness to pay for it posed a potential barrier to the program’s sustainability. A favorable digital health perception within the organization and shared values helped drive adoption, but challenges such as insufficient internal communication, resource constraints, and workflow misalignments hindered the successful integration of the app into routine care. Furthermore, missed opportunities by specialists to introduce the app, coupled with a lack of structured reflection and evaluation processes, contributed to lower adoption and sustained use rates. These findings highlight the necessity for unified commitment at both individual and institutional levels to fully harness the potential of digital health technologies.

This study revealed that factors across all CFIR domains collectively influenced both the app’s acceptability and the clinic’s implementation readiness. In digital health, acceptability is a fundamental yet complex concept, broadly defined as users’ attitudes toward, willingness to engage with, actual use of, or satisfaction with an innovation [36].

The adoption of innovation extends beyond simply introducing new technology; it also requires shifts in individual and group attitudes and behaviors [37]. Behavioral theories suggest that behavior change is a multifaceted process shaped by both individual readiness and external influences [38-40]. For instance, Rogers’ Diffusion of Innovations Theory explains that individuals adopt new technologies at different rates, with relative advantage, the perceived superiority of innovation over existing alternatives, being a key determinant of adoption [39]. Rogers further argues that while relative advantage can be measured in economic terms, factors such as social prestige, convenience, and satisfaction also play a critical role in the adoption process. According to his theory, the greater the perceived relative advantage, the faster an innovation is adopted [39]. Our findings align with this framework, as we identified several influential factors, both in this study and previous research [39,41-45] that could shape app adoption. These include pre-existing lifestyle habits, ease of integration into daily routines, personality traits (eg, being organized and motivated), and perceived value and satisfaction with the app. These results emphasize the importance of assessing patient needs and motivations prior to introducing digital health interventions including mobile apps. Prioritizing digitally ready, motivated users who are most likely to benefit from the intervention can enhance resource efficiency, as well as improve long-term engagement.

This also reinforces a critical insight for mobile app designers and implementers: achieving 100% adoption within the target population is unlikely, despite being the ideal goal. Furthermore, aligning with existing evidence [46], we found that digital literacy does not necessarily equate to digital readiness, as reflected in the lower adoption and engagement rates among younger participants. This challenges commonly accepted assumptions that younger, tech-savvy individuals are inherently more likely to engage with mobile health apps. Instead, our findings suggest that targeted implementation strategies, tailored to the specific characteristics and motivations of the intended users, may be more effective than relying on general demographic trends.

In addition, our results highlighted several app-related features that negatively influenced its acceptability, including limited interoperability and the need for manual data entry. These limiting factors are similarly seen to result in poor acceptability when health technology is perceived to be “more trouble than it is worth” [47,48], too time-consuming [49], unreliable [47], and does not fit the user’s workflow or routine [50,51]. Prior research has highlighted the potential benefit of the inclusion of features such as professional or peer interaction [41,52,53], passive monitoring over manual input [54], and advanced feedback mechanisms [40] to enhance app adoption and sustained use. To improve implementation, future iterations of the app should prioritize seamless integration with existing clinical systems, as well as reduce manual workload. Additionally, incorporating automated data capture and real-time feedback tailored to individual patient needs could further support engagement and long-term adherence. Considering these features during the app development stage or obtaining inputs from the specific target population on their preferences for these features could prove advantageous [55].

Readiness for implementation is a multifaceted construct in implementation science, encompassing stakeholders’ willingness and capacity to adopt new practices [56-58]. Commonly acknowledged individual factors influencing readiness include attitudes, commitment, and self-efficacy [56], and organizational-level factors influencing readiness include climate and resources [59]. At an individual level, the clinic’s readiness was limited by mixed attitudes toward the evidence for short-term app-based exercise programs in atrial fibrillation, which, in turn, influenced clinicians’ commitment to introducing the app during time-restricted consultations with patients. This also impacted uptake as patients were more inclined to adopt and engage with the app when their specialist endorsed it [60-62].

Furthermore, only one clinician in this study regularly used patient-entered app data for clinical management. Clinicians have been shown to adopt new technology when its benefits to their professional roles are evident [47,48]. Limited app introduction and use of collected data highlight a clear gap in clinician buy-in in this study. Gaining buy-in from clinicians, particularly around the evidence base and its value-add to clinical management, is a critical exercise in any health technology implementation [63]. This challenge of gaining clinician buy-in can be a complicated endeavor with an absence of consensus on what defines effectiveness in digital health [64]. At present, lack of scientific validation is a common problem among mobile health apps [65,66].

Engaging health care professionals and consumers early in the design process can help ensure that the intervention (both the app and its implementation) is seamlessly integrated into clinical workflows, its outputs are effectively used to enhance patient care, and it delivers meaningful value to patients. Additionally, establishing a clear evidence base through pilot testing, iterative feedback, and real-world feasibility studies should be prioritized from the outset to build confidence in the app’s effectiveness. These factors are critical for improving clinician buy-in and facilitating successful implementation.

At an organizational level, limited resource availability, poor internal communication, no internal champion or organizational mandate, the absence of ongoing monitoring and evaluation, and the lack of a long-term funding model limited the effective integration of the app. Available literature highlights that proactive and strategic planning during the pre and peri-implementation phases can help mitigate some of these challenges [51,67,68]. Establishing clear leadership roles within the implementation team, securing sustainable funding early in the process, and integrating structured monitoring frameworks to regularly track progress and address emerging challenges are crucial for the successful integration of digital health interventions, including mobile apps, into routine practice. Proactive planning and continuous stakeholder engagement can further enhance adoption and long-term sustainability.

Prior literature has highlighted significant inertia and resistance to new technology adoption within health care settings [67,69] and these factors contribute to slower adoption of new technologies in health relative to many other industries [47,68,69]. We observed the scenario in this study setting to be different. The clinic demonstrated a proactive approach to adopting digital technology, a tendency likely influenced by the positive attitudes of the staff, particularly the leadership [48]. However, despite this inclination toward digital technology in the current study setting, the implementation process remains challenging. This learning highlights the importance of strategic planning even in environments with a strong predisposition toward digital innovation.

Another key finding from this study is the absence of a clear funding model for the mobile app, which, as the literature suggests, poses significant challenges for its long-term sustainability [67,68]. Funding through patient fees, while straightforward, may exacerbate health inequities if patients are unable or unwilling to pay as indicated by our participants. External funding from bodies like health insurers is a viable alternative, but it necessitates a robust evidence base demonstrating the app’s clinical- and cost-effectiveness. To ensure long-term success and equity in digital health tool implementation, health care organizations must prioritize resolving these challenges during the implementation planning stage.

While this study was conducted in a single private clinic, we acknowledge that findings may differ in public hospitals, or nonmetropolitan settings, where resources and infrastructure may be more limited, internet connectivity less reliable, and clinical workload more demanding, potentially creating additional barriers to digital health adoption. However, our key findings related to the app itself, user engagement, and clinic-level implementation factors remain highly relevant across different health care contexts. Additionally, conducting this study in a private clinic may have influenced aspects such as clinician autonomy in adopting digital tools, patient demographics with potentially higher digital literacy, and financial flexibility which could differ in publicly funded or nonmetropolitan settings. Factors like financial constraints and scalability challenges in larger health systems may also shape how similar interventions are implemented more broadly. Future research should explore these contextual differences to better inform implementation strategies across diverse health care environments.

We interviewed approximately one-quarter of patients, 18 out of 68 total patients who signed up for a cardiac rehabilitation care plan until June 2024 since project inception, with varying app use patterns. Unfortunately, despite considerable attempts by the research team to secure access to patient-level app use data to quantitatively explore use patterns or analyze associations between use and participant characteristics, this was not possible. Data on login frequency and feature usage were also unavailable, limiting our analysis to qualitative methods due to reasons beyond our control.

Using an implementation science framework enhanced the systematic examination of the qualitative data, by adding an additional lens through which to consider the code categorization and theme generation. This process pressed the authors not only to consider the data through their normal inductive process but also which parts of the implementation were affected, and in what ways.

The broader implications of our findings may be limited due to several factors. The small sample size restricts the extent to which our results can be applied to wider populations, and the low number of patients who fully completed the cardiac rehabilitation program limits our ability to draw definitive conclusions about long-term patient engagement. However, the challenge of poor patient engagement with cardiac rehabilitation programs [70-72], as well as in the adoption and sustained use of mobile health apps for disease navigation [73,74], aligns with results observed in broader research. Future research with larger, more diverse samples is needed to further explore these findings and assess their applicability across different health care settings.

Additionally, the study was conducted in a private cardiology clinic, which may limit the generalizability to dissimilar settings, particularly to public hospital settings or rural and low-resource environments. However, many of the challenges and facilitators identified such as the need for clinician buy-in, streamlined workflows, and patient engagement strategies are relevant across various health care contexts. In public hospitals, resource constraints and higher patient volumes may necessitate additional support mechanisms, while rural and low-resource settings may require adaptations to address differing levels of digital maturity and access to care. Future research may seek to examine how these findings can be adapted and scaled across diverse clinical environments. Nonetheless, this study highlights the importance of integrating contextual analysis and proactive strategic planning to enhance the adaptability of such interventions.

The expansion of digital technology in routine patient care is expected to accelerate globally. However, it is essential to recognize that digital technology serves only as one means to achieve broader objectives. Its success in reaching these objectives depends on various dynamic and contextual factors within the health system and care teams, as well as the target population. Unless these multifaceted and intertwined factors are assessed and addressed before and during the implementation, health technology integrations are more likely to fail than to succeed [75,76]. Therefore, we advocate for a proactive approach to assessing contextual challenges and strategically planning for their mitigation, as discussed in this study.