Governance fragmentation and sociotechnical misalignment in hybrid telemedicine care pathways: A qualitative study

Study design

This qualitative study employed reflexive thematic analysis. ²³ Individual in-depth interviews (IDIs) and focus group discussions (FGDs) were conducted to examine governance fragmentation and process-level vulnerabilities within a hybrid telemedicine care pathway. The study adhered to the Consolidated Criteria for Reporting Qualitative Research (COREQ). ²⁴ The completed checklist is provided in Supplementary Material A.

The study was informed by a sociotechnical implementation perspective conceptually aligned with the NASSS framework, ¹⁰ which was used not as a coding framework but as an interpretive lens during data analysis and the subsequent discussion. Initial coding and theme development were conducted inductively, allowing patterns of process-level vulnerabilities to emerge from the data. These themes were then interpreted in relation to NASSS domains to support a sociotechnical understanding of cross-level misalignment across sequential stages of care.

Setting and context

The study was conducted in South Korea, where telemedicine operates under a government-led pilot program and is transitioning toward formal institutionalization. The regulatory environment is characterized by evolving legal arrangements, digital platform involvement, and multi-actor governance structures, providing an ideal opportunity to examine hybrid telemedicine pathways under conditions of regulatory transition and institutional complexity.

Participants and sampling

Participants were purposively recruited using a maximum variation sampling strategy to capture diverse experiential perspectives across the telemedicine care pathway. ²⁵ Participants included physicians and patients, as they are directly involved in key stages of telemedicine delivery. The patient group included individuals with disabilities to highlight accessibility-related vulnerabilities. Participant characteristics are summarized in Table 1.

OPEN IN VIEWER

Table 1.

Characteristics of study participants across stakeholder groups.

Characteristic	Physicians (n=5)	Non-disabled patients (n=5)	Patients with disabilities (n=3)
Sex	Male (3), Female (2)	Male (2), Female (3)	Male (3)
Age range (years)	30s-50s	30s-60s	40s-60s
Primary clinical specialties (multiple responses allowed) a	Internal medicine (3); Dermatology (2); Family medicine (1); Hospice & palliative care (1); Home-based care (1); Orthopedic surgery (1)	NA	NA
Primary health condition (patients only)	NA	Chronic conditions (n = 3: hypertension, dyslipidemia, osteoporosis)	Spinal cord injury (mobility-related disability)
Institution type (physicians only) a	Tertiary hospital (1); General hospital (1); Clinic (3)	NA	NA
Clinical experience (physicians only)	<10 years (1); 10-20 years (3); ≥ 30 years (1)	NA	NA
Data collection method	Individual in-depth interviews	Focus group discussions	Focus group discussions

Eligibility criteria were defined a priori. Physicians were required to have direct experience providing telemedicine services during the national pilot program. Patients were eligible if they had at least one prior telemedicine consultation within the pilot program, while individuals without telemedicine experience were excluded. For focus group participation, eligibility was further limited to individuals able to engage in face-to-face group discussion.

Five physicians who had provided telemedicine services during a national pilot program participated in IDIs. They were recruited using purposive and snowball sampling to ensure variation in institutional settings and clinical specialties. Initial participants were identified through academic societies and professional networks engaged in telemedicine research and guideline development and were approached via professional networks and email contact. Additional participants were recruited through peer referral.

Two FGDs were conducted with telemedicine users, one each with non-disabled patients (n=5) and one group of patients with disabilities (n=3). Non-disabled participants were recruited through a professional research agency using a structured screening questionnaire to verify telemedicine experience, and were invited via telephone.

Participants with disabilities were recruited in collaboration with a national disability organization. Initial contact and recruitment were conducted by telephone. Given the FGD format and communication requirements, only individuals able to participate in face-to-face group discussion were included. The final group comprised three participants with spinal cord injuries. None of the participants withdrew after enrollment.

Data collection

IDIs and FGDs were conducted face-to-face by members of the research team (Oh IH, Sul AR, Ju J, and Seo S), comprising two female and two male researchers trained in qualitative research and health policy. Each session was facilitated by one researcher, with a second researcher present as co-facilitator. The remaining two researchers observed from a separate room and recorded field notes during and immediately after each session to document contextual observations and non-verbal dynamics. No non-participants were present in the IDI or FGD room. All sessions were conducted in private meeting rooms at a research facility.

Participants were informed in advance of the study purpose and the institutional affiliations of the research team. No prior personal relationships existed between researchers and participants before recruitment.

Physician IDIs were conducted between September 2 and 4, 2025, and patient FGDs were held on September 3 and 9, 2025. Sessions lasted approximately 90–120 minutes and were audio-recorded with participants’ consent.

Semi-structured guides for IDIs and FGDs were developed by the research team, informed by the sociotechnical perspective. Although formal validity testing is not typically applied in qualitative data collection research, these guides were reviewed and refined through team discussion and expert review prior to data collection. The initial data collection phase functioned as an iterative refinement process, allowing minor adjustments to the guides to improve the clarity, flow, and relevance of questions as data collection progressed. The full guides are provided in Supplementary Material B. Repeat interviews were not conducted.

For the FGD, specific moderation strategies were employed to mitigate potential group dynamics. Given the heterogeneous composition of the group, the moderator actively facilitated balanced participation by encouraging turn-taking and explicitly inviting input from quieter participants. Probing questions were directed to individuals to elicit diverse and potentially divergent perspectives. Efforts were made to not only maintain moderator neutrality but also avoid reinforcing dominant viewpoints. During analysis, data from the FGD were coded at the level of individual contributions, and divergent or minority views were retained as meaningful data rather than being subsumed under dominant themes. Audio recordings were transcribed verbatim and anonymized prior to analysis. Transcripts were not returned to participants for comments or corrections.

Data analysis

Data were analyzed using reflexive thematic analysis following the approach of Braun and Clarke. ²³ Coding was conducted manually and inductively to allow themes to emerge from the data, using standard office software (e.g., Microsoft Word and Excel); no specialized qualitative data analysis software (e.g., NVivo or Atlas. ti) was used. This approach aligns with reflexive thematic analysis, which emphasizes interpretive engagement with the data rather than reliance on software-assisted coding. Initial line-by-line coding was independently undertaken by multiple researchers to generate codes reflecting process-level vulnerabilities and sociotechnical interactions.

The codes were then compared and grouped into candidate themes, forming an analytic framework that captured misalignment across technological, organizational, workflow, and governance domains. The derived themes were refined through iterative team discussions emphasizing conceptual coherence rather than statistical agreement.

Despite the involvement of multiple researchers in coding, the analytic process did not aim to achieve inter-coder reliability or coding consensus. Instead, coding was treated as an interpretive and iterative process, with coding differences used to deepen analytic insight through reflexive discussion.

Although the semi-structured guides for IDIs and FGDs followed the operational phases of telemedicine, the final thematic structure was analytically reorganized into a stage-based framework developed in this study to support process-oriented interpretation, as reflected in the mapping of vulnerabilities across stages (Table 2) and the cross-domain synthesis (Table 3).

OPEN IN VIEWER

Table 2.

Process-level vulnerabilities across stages of the hybrid telemedicine pathway.

Stage	Responsibility ambiguity	Workflow discontinuity	Technological constraint	Governance fragmentation
Appointment	Variable interpretation of eligibility criteria	Multiple, non-integrated booking channels	Platform navigation barriers	Absence of unified triage authority
Identity verification	Ambiguous authentication ownership	Inconsistent verification procedures	Limited authentication mechanisms	Overlapping institutional-platform accountability
Data submission	Clinical-platform responsibility gap	Redundant documentation processes	Overly rigid digital intake forms	Incomplete EHR integration
Teleconsultation	Ambiguity in modality selection	Time compression during encounters	Audio-only assessment limitations	Platform-determined consultation rules
Prescription	Unclear liability boundaries	Fragmented prescription transmission workflow	Inconsistent platform-institution interfaces	Regulatory-platform misalignment
Medication access and follow-up	Undefined follow-up ownership	Weak cross-stage coordination	Limited pharmacy system linkage	Absence of standardized escalation pathways

OPEN IN VIEWER

Table 3.

System-level responses to cross-stage vulnerabilities in hybrid telemedicine.

Stage	Key vulnerability	System-level action	Lead stakeholders
Appointment	Variable interpretation of eligibility criteria	Develop standardized triage criteria and unified scheduling systems	Healthcare organizations; Regulators
Identity verification	Ambiguous allocation of authentication responsibility	Establish consistent authentication standards across institutions and platforms	Healthcare organizations; Platforms; Regulators
Data submission	Incomplete and non-integrated pre-consultation data	Integrate digital intake forms with EHR systems and incorporate accessibility-related fields	Healthcare organizations (EHR); Platforms (form design); Regulators (data standards)
Teleconsultation	Modality misalignment and time compression	Implement complexity-based modality triage and protected consultation time for high-risk encounters	Healthcare organizations (workflow design); Platforms (modality constraints)
Prescription	Fragmented transmission processes and unclear liability	Standardize digital prescription workflows and clarify accountability boundaries	Healthcare organizations; Platforms; Regulators
Medication access and follow-up	Undefined follow-up ownership and escalation criteria	Establish structured follow-up protocols and standardized escalation pathways within hybrid models	Healthcare organizations; Regulators

EHR, electronic health records.

Analytic rigor was supported through reflexive memoing, systematic documentation of analytic decisions, and ongoing team dialogue to enhance credibility, dependability, and confirmability in line with established trustworthiness criteria. ²²

Thematic saturation was assessed iteratively. After seven IDI and FGD sessions, no substantively new conceptual categories or governance-related vulnerabilities emerged. The final session confirmed thematic redundancy and analytic depth across stakeholder groups. Saturation was conceptualized as meaning saturation and information power, focusing on conceptual depth rather than on a predefined numerical threshold,^26,27 emphasizing the adequacy of analytic insight and thematic richness rather than the frequency of code emergence. Participant checking of findings was not conducted.

The final themes represent analytically derived patterns of sociotechnical vulnerability across stages of care and are organized into four domains: responsibility ambiguity, workflow discontinuity, technological constraint, and governance fragmentation.

Reflexivity

The research team comprised faculty members and health policy researchers affiliated with academic and public research institutions at the time of the study. All researchers involved in data collection had doctoral-level training in health policy, digital health, and qualitative research. Several researchers had prior professional involvement in telemedicine policy research, providing contextual familiarity and the potential to shape analytic interpretation.

Reflexive discussions were conducted throughout data collection and analysis to critically examine how prior assumptions regarding governance fragmentation, implementation challenges, and digital equity might influence coding and theme development. This process informed analytic decisions and was used to enhance transparency, credibility, and confirmability. ²²

Ethical considerations

Ethical approval was obtained from the Institutional Review Board of the National Evidence-based Healthcare Collaborating Agency (NECAIRB 25-019). All participants provided written informed consent prior to participation, including consent for audio-recording. Identifying information was removed from transcripts prior to analysis, and data were stored securely in password-protected systems accessible only to the research team. To further protect participant confidentiality given the small group sizes, quotations presented in the Results are attributed to participant groups (physician, non-disabled patient, or patient with disability) without individual identifiers.

Participant characteristics

As summarized in Table 1, participating physicians represented diverse clinical specialties, institutional settings, and levels of telemedicine experience within the national pilot program. Patient participants varied in age, sex, and health status. Most non-disabled patients reported chronic conditions requiring ongoing management, indicating that telemedicine use was embedded within longitudinal care rather than limited to episodic consultations. Participants with disabilities had established care needs and prior telemedicine experience, enabling examination of accessibility-related vulnerabilities within hybrid care delivery.

Overview of system-level vulnerabilities

The analysis identified systemic vulnerabilities across all stages of the hybrid telemedicine care pathway that did not reflect isolated technical failures but emerged from sociotechnical misalignment across governance arrangements, clinical workflows, platform design, and regulatory oversight.

Six sequential stages were examined: appointment, identity verification, pre-consultation data submission, teleconsultation, prescription, and medication access and follow-up. Across these stages, vulnerabilities clustered into four analytically derived domains, which represent the core themes of this study: (1) responsibility ambiguity, which refers to unclear allocation of decision-making authority across actors; (2) workflow discontinuity, which denotes breakdowns across sequential care processes; (3) technological constraint, which reflects structural limits imposed by platform design and interface architecture; and (4) governance fragmentation, which captures misalignment across institutional, regulatory, and platform actors.

Participants described how these vulnerabilities accumulated across stages:

“It’s not just one issue—small problems at each step build up and make the whole process feel unreliable.” (Physician)

Table 2 synthesizes these findings into a stage-by-domain diagnostic framework, illustrating how distinct forms of sociotechnical misalignment arise at specific interfaces along the telemedicine pathway. Rather than isolated issues, these vulnerabilities are embedded in routine care processes and distributed across system components.

Table 3 translates these diagnostic insights into a system-level response framework by linking key vulnerabilities to corresponding actions and responsible stakeholders. This framework outlines an operational approach to align governance structures, clinical workflows, and platform design, highlighting the role of coordinated stakeholder action in addressing cross-stage vulnerabilities in hybrid telemedicine systems.

Importantly, vulnerabilities did not operate independently but accumulated across stages, reinforcing uncertainty and reducing confidence in telemedicine as a continuous mode of care rather than as an isolated digital encounter. Together, these tables provide a pathway-oriented conceptual and practical framework that can inform both the evaluation and implementation of hybrid telemedicine systems.

To summarize these findings, we developed a conceptual diagram (Figure 1) that maps the distribution and emergence of vulnerabilities across sequential stages of the hybrid telemedicine pathway.OPEN IN VIEWER

Stage 1: Appointment – Access variability and responsibility ambiguity

During the initial engagement, participants described substantial variability in how telemedicine services were accessed and coordinated. Physicians operated across multiple booking channels—including institutional systems, commercial platforms, and direct requests—without a unified scheduling framework:

“Depending on the platform or hospital, the booking system is completely different, so there’s no unified way to manage appointments.” (Physician)

Patients encountered inconsistent entry pathways, contributing to uncertainty regarding their eligibility and timing:

“I wasn’t sure if I could use telemedicine or how to start—it depends on where you hear about it.” (Non-disabled patient)

Divergent sources of telemedicine awareness shaped expectations. Physicians described structured familiarity with policy and procedural guidance, whereas patient awareness was often informed by fragmented sources such as media or platform advertisements. This divergence contributed to uneven expectations regarding the scope of care and limitations of remote consultation, particularly among first-time users. Scheduling thus functioned as an early trust-forming stage within the hybrid pathway, reflecting variability in eligibility interpretation and fragmented booking structures.

Stage 2: Identity verification – Procedural inconsistency and safety concerns

Identity verification practices varied across institutions and platforms. Some settings relied on minimal authentication procedures, raising concerns regarding impersonation risk and misidentification:

“Sometimes identity verification is very basic, and it makes me uneasy about whether the patient is correctly identified.” (Physician)

Responsibility for verification was perceived as ambiguously distributed between providers and platforms:

“It’s unclear whether the hospital or the platform should take responsibility for verifying the patient’s identity.” (Physician)

Participants also noted inconsistencies in verification procedures:

“Sometimes the identity check feels minimal while, at other times, it’s more thorough—it’s not consistent.” (Non-disabled patient)

Participants linked inconsistent authentication standards to broader concerns regarding procedural reliability and patient safety, especially in contexts involving prescriptions or sensitive health information. Such inconsistencies reflect underlying ambiguity in the allocation of responsibilities and the variability in verification processes across platforms.

Stage 3: Pre-consultation data submission – Fragmented information flow

Pre-consultation data submission processes were described as insufficiently integrated with electronic health records. Physicians noted that digital questionnaires were frequently overly simplified and lacked clinically meaningful detail. This led to redundant information verification during consultations, undermining intended workflow efficiencies:

“The questionnaire is too basic; so I still have to ask everything again during the consultation.” (Physician)

Patients reported difficulty navigating structured digital forms, particularly when medical terminology was used:

“The questions are too limited, and it’s hard to explain my condition properly.” (Non-disabled patient)

Patients with disabilities emphasized that functional status, mobility constraints, and contextual care needs were rarely captured adequately:

“There’s no place to explain disability-related issues, which are important for my care.” (Patient with disability)

These documentation gaps were linked to concerns about clinical safety, decision-making reliability, and alignment between submitted information and actual care needs, corresponding to workflow discontinuity, technological constraints, and incomplete system integration.

Stage 4: Teleconsultation – Interactional and diagnostic constraints

Although teleconsultation improved convenience and reduced travel burden, physicians reported reduced diagnostic confidence, particularly in audio-only encounters or when contextual cues were limited:

“Without seeing the patient directly, it’s harder to make accurate judgments, especially in complex cases.” (Physician)

Consultation modality was often predetermined by platform configuration rather than by clinical complexity:

“Whether it’s video or phone isn’t always based on the patient’s condition.” (Physician)

Both physicians and patients characterized teleconsultations as compressed and standardized, shaped by system constraints rather than by individualized care needs:

“It feels shorter and more rushed than in-person visits.” (Non-disabled patient)

While efficiency was acknowledged, participants described tension between convenience and relational depth, especially when managing complex or evolving clinical conditions. These findings reflect modality ambiguity, time compression, and platform-driven constraints.

Stage 5: Prescription – Regulatory and platform misalignment

Prescription transmission revealed tensions between digital efficiency and regulatory safeguards. Physicians reported fragmented workflows linking telemedicine platforms, healthcare institutions, and pharmacies:

“Different systems don’t communicate well with each other.” (Physician)

Ambiguity in liability distribution between clinical decision making and platform-mediated prescription processes was frequently noted:

“If something goes wrong, it’s not clear whether the responsibility lies with the doctor, the platform, or the pharmacy.” (Physician)

Participants reported uncertainty regarding medication processes:

“Sometimes the prescription doesn’t reach the pharmacy smoothly.” (Non-disabled patient)

Breakdowns at this stage were described as undermining confidence in telemedicine as a reliable modality of care, reflecting fragmented workflows, interface limitations, and regulatory–platform misalignment.

Stage 6: Medication access and follow-up – Continuity gaps

Medication access and follow-up processes lacked standardized coordination mechanisms, with patients reporting uncertainty regarding dispensing locations and monitoring responsibilities:

“After the consultation, I wasn’t sure who to contact if something changed.” (Non-disabled patient)

These vulnerabilities were amplified among individuals with mobility limitations or reduced pharmacy access:

“Getting the medication itself can be difficult, especially when mobility is limited.” (Patient with disability)

During transition to in-person care, escalation criteria were often unclear. Physicians and patients described ambiguity regarding responsibility for monitoring and referral decisions:

“There’s no clear guideline on when to switch to face-to-face care.” (Physician)

This stage proved pivotal in determining whether telemedicine functioned as an integrated component of a hybrid care continuum or as an isolated, episodic interaction, reflecting gaps in follow-up ownership, coordination, and escalation pathways.

Cross-stage finding: Governance fragmentation

Across all stages, vulnerabilities were amplified by fragmented governance structures, with platforms, clinicians, regulators, and patients operating within partially overlapping yet insufficiently aligned responsibility frameworks:

“There are multiple actors involved, but no clear coordination between them.” (Physician)

Rather than discrete operational failures, the findings highlight the systemic uncertainty generated by misaligned institutional arrangements. Governance fragmentation functioned as a cross-cutting condition shaping responsibility allocation, workflow coordination, technological design constraints, and continuity of care throughout the hybrid telemedicine pathway.

Telemedicine as a longitudinal sociotechnical pathway

This study identifies telemedicine as a longitudinal, multi-stage sociotechnical pathway rather than as a series of isolated digital encounters. Implementation challenges arose less from resistance to digital technologies than from how care processes were structured, coordinated, and governed across sequential stages. Breakdowns were consistently observed at interfaces between scheduling, pre-visit preparation, consultation, prescribing, and follow-up, highlighting the embedded fragmentation within routine care processes rather than that arising from discrete technical failures. Similar patterns of responsibility ambiguity and workflow fragmentation have been observed in chronic care and telemonitoring contexts, ²⁸ further supporting this pathway-oriented conceptualization and suggesting that integration unfolds across interconnected organizational processes rather than through discrete clinical events.

Importantly, this pathway-oriented perspective extends prior research that has typically examined telemedicine challenges within single domains—such as technology performance, patient experience, or policy constraints—by demonstrating how vulnerabilities accumulate across sequential stages of care. This perspective advances pathway-oriented evaluation beyond single-domain analyses, aligning with broader digital health literature emphasizing that healthcare transformation is driven by not only technological advancement but also the integration of technological, organizational, and social-ethical dimensions as interconnected domains.^29,30

By mapping vulnerabilities to specific pathway interfaces, this study advances a pathway-oriented sociotechnical perspective that specifies where and how governance misalignment manifests within routine hybrid care processes.

Telemedicine challenges as sociotechnical misalignment

Consistent with implementation research, sustained telemedicine use seemingly depends more on organizational alignment, workflow compatibility, and institutional coordination than on clinician attitudes alone.^6,9 While clinician perceptions were generally favorable, vulnerabilities emerged from interactions among platform design, workflow structure, and governance arrangements.

Extending NASSS-based evaluations, these findings reveal how adoption, scale-up, and sustainability challenges arise from interacting complexities across technological, organizational, and regulatory domains. ³¹ Telemedicine implementation therefore reflects coordination failures across domains rather than deficits in technology acceptance.

This interpretation is consistent with prior studies, which reveal that digital transformation in healthcare is contingent on organizational capability and leadership alignment, particularly in resource-constrained settings, where institutional readiness and leadership support shape implementation outcomes. ³² At the institutional level, leadership perceptions of digital capability have been shown to directly influence the quality of patient-centered care, with leadership support for digital adoption and availability of trained ICT personnel identified as the most significant contributing factors in lower-resource hospital settings. These findings suggest that sociotechnical misalignment is not solely a technological issue but also reflects gaps in cross-level governance and organizational coordination.

These patterns should be interpreted in the context of the South Korean healthcare system, which is characterized by high accessibility under a universal national health insurance scheme alongside structural concentration of patients and resources in large tertiary hospitals. ²

In particular, prior studies have documented regional differences in not only healthcare access and utilization but also the concentration of specialized care in major hospitals, which may contribute to the uneven access pathways and patient flow patterns observed in telemedicine settings. ²

Pre-visit preparation and information continuity

Insufficient pre-visit preparation emerged as a critical determinant of downstream care quality. Both physicians and patients described limitations in the completeness and usability of pre-consultation information, echoing evidence that inadequate pre-consultation information increases clarification burdens and contributes to diagnostic uncertainty during time-limited encounters.^18,33

From a sociotechnical perspective, pre-visit preparation represents a structural interface between patient-generated data and clinician interpretation. Structured information exchange and communication maturity appear foundational to digital health system capability and implementation effectiveness. ³⁰

In addition, consumer-oriented research on telehealth services has identified service quality and accessibility as primary determinants of user satisfaction and engagement, with innovation emerging as a key pathway through which service challenges are translated into improved care experiences. ³⁴ Our findings extend this literature by demonstrating that such perceptions are not only user-level experiences but are shaped by structural misalignments in information continuity across care settings.

Accessibility as a cross-cutting system property

Accessibility-related vulnerabilities were evident across all stages, particularly among patients with disabilities. Exclusion risk extended beyond dissatisfaction to the functional inability to engage with digital systems, consistent with quantitative evidence linking frailty and socioeconomic deprivation to reduced telemedicine access. ¹⁵

Qualitative syntheses further indicated that equitable telemedicine delivery requires workflow adaptation, supportive infrastructure, and organizational redesign rather than reliance on technology alone.^16,19,35,36 Our findings underscore the need to conceptualize accessibility not as an individual trait but as a system-level design property intersecting multiple stages of care.

This interpretation is consistent with research on the digital health divide, which demonstrates that disparities in telemedicine access arise from structural factors—including infrastructure, policy environments, and community capacity—rather than from individual-level limitations. ³⁷ By situating accessibility within a pathway-oriented framework, this study highlights how exclusion is produced and reinforced across interconnected stages of care.

Post-consultation workflows, reimbursement, and sustainability

Post-consultation processes—including prescription transmission, medication access, and follow-up—represented areas of heightened vulnerability. Although meta-analyses demonstrate that telemedicine can significantly reduce outpatient waiting times, ³⁸ they primarily assess access metrics, overlooking coordination processes beyond the initial consultation.

Weaknesses in post-consultation workflows and reimbursement alignment may constrain the sustainability of telemedicine integration. ¹² Reimbursement structures and payment models shape how telemedicine is positioned within care delivery, influencing its use for follow-up versus first-contact encounters.^12,39

Legal analyses further demonstrate that fragmented accountability structures and ambiguous professional qualification standards can amplify governance-related risks in telemedicine systems. ⁴⁰ Emerging AI-driven capabilities—including predictive analytics, automated triage, and remote monitoring tools—serve as potential pathways to address coordination gaps and extend service reach in resource-constrained settings, although their implementation introduces additional ethical and regulatory considerations that must be managed alongside governance reform. ³⁷ Sustainable telemedicine delivery therefore depends on managing cross-domain complexity and aligning technological, organizational, financial, and regulatory domains over time. ³¹

Implications for pathway-oriented evaluation frameworks

Recent evaluation frameworks emphasize multi-domain, stakeholder-informed assessment in telehealth, with some incorporating developmental staging approaches.^5,41 A Delphi-based telemedicine implementation framework similarly identifies multi-level constructs spanning clinical, technological, regulatory, economic, and monitoring domains. ⁴²

However, many frameworks lack telehealth-specific operationalization and provide limited guidance on how implementation and equity challenges unfold in routine care settings,^36,43 while evaluation models often remain fragmented or narrowly scoped. ⁴⁴ By contrast, this study integrates multiple domains—technological, organizational, experiential, and policy-related—within a single pathway-oriented framework, complementing prior research on consumer experience, organizational leadership, and digital health equity by demonstrating how these dimensions interact dynamically rather than in isolation across sequential stages of care. This has direct implications for policy design, platform development, and clinical workflow integration at each stage of the care pathway.^32,34,37

Limitations

This study has several limitations. First, it did not include hospital administrators or other meso-level actors, which limited direct examination of organizational governance and decision-making processes. This reflects the study’s intentional focus on the interaction between micro-level clinical practice and macro-level policy contexts, aimed at identifying process-level vulnerabilities in telemedicine care pathways. Although institutional factors were indirectly captured through physician accounts, future studies should incorporate meso-level stakeholders to provide a more comprehensive understanding of governance fragmentation across system levels.

Second, the sample size was modest and participants were recruited through professional networks and a research agency using prior telemedicine experience as a screening criterion. This approach may have introduced selection bias toward individuals with greater familiarity or engagement with telemedicine, potentially underrepresenting the perspectives of those with limited access or lower utilization. In addition, the inclusion of only a small number of patients with disabilities limits subgroup-specific insights, although these participants were included as information-rich cases to identify accessibility-related vulnerabilities. Future research should employ purposive sampling strategies that deliberately recruit underrepresented populations—such as individuals with limited digital access and those across a broader range of disability types—and adopt larger, multi-site designs to strengthen the transferability of findings.

Third, patient data were collected through two FGDs with heterogeneous participant composition, which may have introduced group dynamic effects such as the dominance of certain participants or convergence of views. Although moderation strategies were applied and individual-level analysis was conducted, some influence of group interaction is inevitable. Future research could incorporate complementary individual interviews or stratified focus groups with more homogeneous participant characteristics to reduce group dynamic effects and enrich the depth and diversity of perspectives. In addition, participant checking (member checking) was not conducted, which may limit the extent to which the interpretations were validated by participants.

Fourth, variability across telemedicine platforms and institutional workflows was not examined as an independent analytic dimension. Given that participants reported heterogeneous experiences across platforms, differences in platform design, interface functionality, and system integration may have contributed to the observed workflow fragmentation and access variability. Future studies should systematically compare platform-level characteristics and their impact on care processes.

Finally, as is common in qualitative research, the findings are not intended for statistical generalization and reflect a specific institutional and regulatory context. In particular, the governance challenges identified in this study are closely tied to South Korea’s evolving telemedicine policy environment and hybrid implementation model. Caution is therefore warranted when transferring these findings to other national contexts with different regulatory environments, healthcare infrastructure, and platform governance models. Nevertheless, the pathway-oriented sociotechnical framework may offer analytic relevance to telemedicine settings with similar hybrid care structures, consistent with prior studies.^6,18,36,45