As part of a national mixed methods evaluation examining the implementation and impact of VA’s TEC program, we conducted in-depth interviews with clinicians from TEC programs. The qualitative component was designed to elicit clinicians’ experiences delivering emergency care through TEC, identify perceived benefits and challenges of the model, and inform program improvement. We employed a qualitative descriptive approach [20] to analyze interview transcripts, using constant comparison [21] to iteratively organize, refine, and compare and contrast data across transcripts.

We performed a descriptive qualitative analysis and used summation as our analytic goal. We used summation as our analytic goal because this applied evaluation sought a descriptive synthesis rather than theoretical development. The 15 interviews across 4 programs produced detailed accounts across all major interview domains and sufficient recurrence and variation to support the 4 descriptive themes reported. We therefore judged the sample adequate for these bounded aims, although not sufficient to establish national representation, compare static and distributed staffing models, or capture perspectives across all 18 programs. While researchers have traditionally relied on saturation as an analytic benchmark, many are increasingly finding it to be ill-suited for more applied, health services–oriented research [21]. Summation-driven studies are more modest in scope than saturation-based studies that aim for conceptual and theoretical development [21]. Reporting of this qualitative study followed the COREQ (Consolidated Criteria for Reporting Qualitative Research) guidelines, with the completed checklist included in Checklist 1.

Interviews were conducted between February 2025 and June 2025. We employed a multistage purposeful sampling strategy [22]. In the first stage, we purposefully selected 4 of 18 TEC programs to achieve maximum variation across TEC program type, tenure, geography, and veteran population characteristics (Table 1). In the second stage, the National Emergency Medicine Office asked leads at each of the 4 selected programs to provide the evaluation team with a complete contact list of all TEC staff, rather than to nominate specific clinicians or apply selection criteria. The evaluation team then contacted providers directly via email and Microsoft Teams messages to invite participation in one-on-one interviews. Because leads served as intermediaries in identifying potential participants, the completeness of these lists could not be verified. Of the 16 clinicians contacted, 15 agreed to voluntarily participate in an interview.

The interview guide was developed by our interdisciplinary evaluation team (AE, JF, CG, JS, and AAV), which included qualitative researchers and emergency medicine physicians, 2 of whom worked clinical TEC shifts, with additional input from VA operational partners (Multimedia Appendix 1). The draft guide was then pilot-tested with an emergency medicine physician who also served as a TEC clinician but was not part of the study sample. The guide was largely retained following the pilot, with minor wording clarifications. Interviews explored clinicians’ experiences delivering TEC, challenges with virtual emergency care, needed improvements to the program, and tools and training needs specific to clinicians. Clinicians also discussed the impact of TEC on their overall work satisfaction and its role in addressing burnout. Two trained female qualitative researchers, with backgrounds in sociology (CG) and social work (AE), conducted all interviews. The research team did not have a prior clinical or supervisory relationship with participants. Participant characteristics were recorded from information volunteered during interviews rather than collected through a standardized form; denominators, therefore, vary by characteristic and are reported in Table 1.

Semistructured interviews were conducted using Microsoft Teams, a secure videoconferencing platform. With participant consent, interviews were audio-recorded and professionally transcribed. Interviewers drafted summary memos following each interview to support generative analysis and facilitate debriefings with the broader evaluation team. These debriefings informed iterative refinements to the interview guide, including the addition of follow-up questions or new probes. This approach is consistent with data-driven forms of qualitative inquiry.

Interview transcripts were analyzed using a qualitative descriptive approach that employed a constant comparative analytic technique to further reduce and synthesize data [23,24]. Though typically associated with grounded theory approaches, constant comparison is a way to continually compare previously analyzed data with new data to look for patterns as well as areas of divergence. Our coding process was abductive, involving both “deductive” codes derived from our interview guide and knowledge about the field, as well as “inductive” codes and categories that were based on immersion in the interview data itself [25]. The codebook was developed collaboratively by the evaluation team, informed by the interview guide’s domains and iterative review of early transcripts. Transcripts were uploaded into ATLAS.ti and coded by a single trained qualitative analyst (CG) using the collaboratively developed codebook. Independent double-coding and interrater reliability statistics were not used [25]. The single-coder approach was selected because of evaluation time constraints and limited opportunities to assess coding consistency. To strengthen analytic rigor, coding decisions and emerging categories were discussed with our interdisciplinary evaluation team during twice-monthly team debriefings, providing an additional layer of collaborative review. Regarding positionality, the primary coder (CG) is a social scientist with more than 10 years of experience conducting VA health services research and nearly 10 years evaluating virtual care. AE has more than 10 years of experience in mixed methods research in acute care, including VA program evaluation, as well as prior clinical experience as a social worker in primary care and trauma center settings. CG and AE do not provide TEC. Two emergency medicine physician members of the analytic team (JF and JS) provide TEC, bringing direct clinical and operational knowledge of the program. All emergency medicine physician team members (JF, JS, and AAV) provide in-person clinical care to veterans in VA EDs. Thus, the analysis combined non-TEC analyst perspectives with insider clinical perspectives.

Following coding, team members (AE, JF, CG, JS, and AAV) reviewed and grouped codes into broader categories and identified higher-level themes based on patterns of repetition and emphasis. Finally, quotes were selected to illustrate each of the major themes and subthemes presented in the Results section.

This quality improvement project was reviewed and designated as nonresearch by the Institutional Review Board of Stanford University. The institutional review board has determined that this project does not meet the definition of research as defined in 45 Code of Federal Regulations (CFR) 46.102(d) or the definition of a clinical investigation as defined in 21 CFR 50.3(c). Despite being categorized as nonresearch, oral informed consent was obtained from all participants prior to their involvement in the study. Permission was also granted to audio-record the interviews. All audio recordings were transcribed, and the transcripts were deidentified to ensure the anonymity of participants. This deidentification step was taken to protect the privacy and confidentiality of the participants.

Additionally, to ensure confidentiality, recruitment and interviews were conducted by the evaluation team, and individual responses were not shared with program leadership. Interview participants were provided with a description of the evaluation’s aims, assured that confidentiality would be maintained, and informed that findings would be reported at an aggregate level rather than identified by location or any other potentially identifiable information. Participants did not receive compensation, as VA staff and providers are not allowed to accept compensation for participation in research or evaluation activities.

Participants included 14 emergency medicine physicians and 1 nurse practitioner, all of whom had formal emergency medicine training and experience (Table 1). The mean interview length was 32 (SD 14) minutes.

Participants described four primary ways that TEC affects the delivery and experience of providing emergency care: (1) clinical decision-making in a virtual environment; (2) development of the provider-veteran relationship; (3) clinician job satisfaction and professional well-being; and (4) challenges of TEC (Table 2).

Emergency medicine clinicians are trained to evaluate, stabilize, and resuscitate patients in a resource-rich environment. In contrast, TEC clinicians described a key component of their jobs as actively working to safely avoid sending veterans to the ED whenever possible and found it rewarding when they could accomplish this. Clinicians reported that they found it rewarding when they could resolve concerns virtually and help veterans avoid an ED visit. This is especially valuable for rural veterans, as highlighted in the following quote:

In traditional ED settings, there are abundant diagnostic resources available, including laboratory testing and imaging, making it easy for clinicians to order these workups even if not strictly necessary. TEC clinicians described their work as requiring more thoughtful consideration of whether in-person evaluation would meaningfully change management. With ready access to the veterans’ electronic health record, clinicians described initially risk-stratifying patients and then communicating their professional recommendations virtually. In addition, though always important, clinicians perceived that shared decision-making was particularly critical for TEC, especially for veterans who strongly wished to avoid going to the ED or those who sought guidance on whether in-person ED evaluation was necessary. Clinicians described collaboratively offering outpatient alternatives to emergency care to honor patients’ care preferences, like in this clinician’s patient care example, which states:

When delivering emergency care virtually, participants described how they adapted their medical evaluation in the absence of a traditional physical exam. To compensate for this absence, video visits rather than phone visits were viewed as critical for making informed decisions. Clinicians perceived that visually observing patients in their home environment was essential for assessing functional status and providing a fuller picture of contextual factors that might be impacting the patient’s health, including mobility, baseline independence, access to caregivers, and the safety of the home setting. Clinicians noted that these observations informed risk stratification and helped them determine whether patients could be safely managed at home or required in-person evaluation. As one provider explained:

Beyond visual assessment, clinicians reported that TEC required greater reliance on patients’ self-knowledge of symptoms and lived experience of chronic conditions. In the absence of vital signs and other objective data typically available during in-person ED visits, providers described engaging more deeply in dialogue to distinguish benign from high-risk presentations. This approach often involved validating patients’ interpretations of their symptoms while applying clinical judgment to determine the need for escalation. As illustrated below, clinicians contrasted these nuanced clinical assessments with protocol-driven triage processes that may default to unnecessary ED referral:

In contrast to the fast-paced nature of the in-person ED, where clinicians multitask, have frequent interruptions, and simultaneously manage multiple high-acuity patients, interviewees described the opportunity to have focused, one-on-one interactions with patients on TEC. Clinicians perceived that this uninterrupted time supported rapport-building, facilitated trust, and enabled more patient-centered conversations than are typically possible in the in-person ED environment. One clinician noted:

Clinicians described how TEC created space to acknowledge veteran-specific experiences and values, particularly around ED use. They reported that many veterans preferred to avoid the ED when possible, and that when they could safely honor this preference through virtual care, it reinforced trust and rapport. Clinicians perceived that veterans who experience posttraumatic stress disorder may find the ED environment stressful, with TEC providing a further benefit to such patients. They also noted that some veterans wish to avoid using traditional ED services because they felt that others who might have more urgent needs should be prioritized, and TEC provided those veterans with an additional option for care. One clinician explained:

Clinicians characterized the positive aspects of providing TEC as personally rewarding in terms of minimizing burnout and providing professional meaning. Additionally, some perceived that having time to assist veterans in navigating a complex health care system was particularly rewarding, with one clinician stating:

Clinicians described veterans’ expressions of gratitude during TEC encounters as a source of professional fulfillment and a further buffer against burnout. They reported that patients expressed appreciation for their TEC visits more often than in the high-stress atmosphere of the ED, and feeling valued reinforced clinicians’ sense of purpose. As one provider explained:

Clinicians also frequently described TEC as a positive change of pace that offered greater schedule flexibility and improved work-life integration. Among clinicians who continued to practice in-person emergency medicine, some described benefits of combining virtual and in-person work, including the ability to maintain procedural skills while periodically working from home. Participants also highlighted benefits of virtual work more generally, including reduced late-night commuting, more time with family, decreased exposure to illness, and fewer in-person distractions. However, these favorable experiences were not universal. Clinicians also described frustration when TEC encounters centered more on administrative tasks, care coordination, or concerns outside the traditional scope of emergency medicine.

Despite the reported benefits of TEC, clinicians also described less favorable aspects of the virtual care model, and perspectives differed regarding some responsibilities. Clinicians noted that TEC encounters frequently involved clinical issues outside the traditional scope of emergency medicine, particularly the evaluation and management of chronic or subacute conditions typically addressed in primary care. Some clinicians valued the opportunity to fill these gaps in primary care access, whereas others viewed these responsibilities as a less optimal use of emergency medicine expertise. For example, a clinician explained:

Clinicians described how both the VA system and the broader health care system are still challenging to navigate within TEC. While some clinicians viewed helping patients navigate a fragmented system as an opportunity to help move care forward, others expressed concern that these responsibilities extended beyond the intended role of emergency medicine–trained clinicians. They perceived that TEC served not only as a means of addressing acute clinical needs but also as a way to help veterans navigate a complex, fragmented system, suggesting that additional supports for veterans may be needed to address these challenges. A clinician explained:

In this example, the clinician reported being able to assist the veteran, but time constraints and other obstacles may prevent some TEC clinicians from offering this additional help. Finally, in some cases, clinicians reported that patients had already experienced system inefficiencies, such as multiple phone transfers or an inability to reach their outpatient clinicians, before reaching TEC. These inefficiencies at times contributed to patient frustration and made TEC encounters feel suboptimal. One interviewee noted:

These experiences suggest that TEC may shift, rather than eliminate, some of the burdens encountered in traditional in-person emergency care, particularly when clinicians are asked to compensate for gaps elsewhere in the health care system.

Our study provides insight into emergency medicine clinicians’ experiences delivering care through VA’s novel TEC program, which positions emergency clinicians upstream of the ED to address urgent concerns in veterans’ home environments. Compared with in-person emergency care, clinicians described how TEC reshapes clinical decision-making, the provider-patient relationship, and clinician well-being. Although many participants perceived benefits, including greater schedule flexibility and more focused patient interactions, others described challenges related to care coordination, work outside the traditional scope of emergency medicine, and system inefficiencies.

Prior VA-based quantitative studies have demonstrated that TEC can resolve acute concerns without ED referral in a substantial proportion of encounters and reduce subsequent ED use without adverse effects on safety outcomes [9,12]. An important feature of TEC is its use of emergency medicine clinicians for virtual evaluation and disposition decisions. Research in other integrated health systems similarly suggests that this upstream, physician-involved virtual approach can safely reduce reliance on in-person emergency care. A Kaiser Permanente study found that physician-directed telephone triage for patients with chest pain was associated with fewer ED referrals and shorter calls than nurse-directed triage, without significant differences in hospital admission or short-term mortality [26]. Our qualitative findings add context to how virtual emergency care is delivered in practice to guide these disposition decisions. Clinicians reported the importance of video-based encounters for assessing functional status, home environment, and available support, noting that these contextual insights supported decisions about when virtual management might be appropriate and when in-person evaluation was warranted. From the patient’s perspective, prior literature has shown that patients also prefer video over phone visits when engaging with clinicians, as it facilitates communication and connection, although preferences may vary by visit type [27-29]. For example, brief or straightforward concerns such as discussing test results may be well suited to phone-based encounters.

The challenges clinicians described in TEC are largely consistent with those encountered during in-person emergency care. As the safety net of the health care system, EDs frequently fill gaps created by limited primary care access, high demand for clinicians, and long waits for outpatient appointments. Patients often present to the ED because they struggle to navigate the system, including coordinating specialty care. Patients’ frustration stemming from system inefficiencies (eg, multiple phone transfers or difficulty reaching outpatient clinicians) that clinicians reported also mirrors common in-person ED experiences, where overcrowding, delays, fragmented communication, and lack of timely follow-up are frequent drivers of dissatisfaction. Notably, clinicians differed in how they viewed addressing these responsibilities in TEC. Some found helping veterans navigate fragmented care professionally rewarding, whereas others viewed care coordination and management of nonacute concerns as extending beyond the intended role of emergency medicine–trained clinicians.

Participants described schedule flexibility and focused patient interactions as benefits of TEC. Among clinicians who also maintained in-person ED practice, some valued combining the 2 settings, noting that in-person work could help preserve procedural skills and professional identity while virtual shifts reduced commuting time and occupational exposure to illness. Prior studies in other specialties report mixed clinician experiences with hybrid practice, including comparable job satisfaction and concerns about reimbursement [30-32]. These accounts suggest that mixed virtual-in-person roles may be attractive to some clinicians, but they should not be interpreted as a uniform preference. Sixty percent of participants in our study were affiliated with static programs (where the clinician primarily provides TEC services with little to no in-person ED shifts), and we did not systematically assess staffing preferences or each clinician’s distribution of virtual and in-person work outside TEC. We therefore cannot compare static and distributed programs (where the clinician primarily provides in-person ED care with occasional TEC shifts) or conclude that a hybrid model is superior. Future work should directly evaluate how staffing configuration relates to clinician satisfaction, skill maintenance, retention, and program capacity. Clinicians perceived that more focused, uninterrupted virtual encounters facilitated more patient-centered relationships; however, patient satisfaction and trust were not directly assessed in this study. Additionally, these favorable perceptions should be interpreted in light of potential selection bias, as clinicians who voluntarily participated may have been more engaged with or supportive of TEC.

Across staffing models, clinicians reported that delivering emergency care through TEC required adapting traditional ED clinical reasoning to a virtual setting, with greater reliance on history and physical exam via video, and shared decision-making in the absence of immediate diagnostics. Clinicians perceived that this model may be particularly useful for veterans who face geographic or logistical barriers to timely in-person acute care. Based on these reported experiences, training and onboarding for virtual care positions should explicitly address clinical decision-making without a traditional physical exam and shared decision-making strategies tailored to virtual care. Given how widespread virtual care options across specialties have become since COVID-19, these competencies could be introduced as early as medical school and reinforced in residency curricula [33-35].

Notably, our participants did not describe the concerns most prominent in prior studies of virtual emergency and urgent care. In a single-site virtual ED, a subset of physicians reported that their emergency medicine skills were underutilized and that patient issues were often unresolved [10], and clinicians delivering direct-to-consumer tele-urgent care described tension arising from patient expectations for prescriptions or testing [11]. Neither pattern appeared in our interviews. Structural differences may contribute to this finding. For instance, TEC encounters are initiated through nurse triage rather than patient self-referral, virtual shifts are incorporated into clinicians’ regular schedules rather than added to them, and clinicians are salaried rather than compensated per encounter. These features may also influence how readily virtual emergency care can be implemented outside the VA. Although physician-led virtual urgent and emergency care has been implemented at academic medical centers and individual hospitals [10,11], nonintegrated systems may have more fragmented access to medical records, ordering, referrals, and follow-up. Similar limitations can occur within the VA when veterans receive care outside the system. Reimbursement and referral pathways may be particularly important, as they may influence not only the feasibility of virtual emergency care but also clinician workload, patient expectations, and the clinical encounter itself.

Several implications emerge for the implementation and expansion of virtual emergency care, more generally, given our findings. Further work is needed to define and evaluate quality and safety metrics, particularly for high-acuity presentations. This is especially important, given that TEC clinicians must adapt traditional emergency medicine decision-making to a setting without a hands-on examination or immediate diagnostic testing. Triage processes should also be evaluated to ensure that emergency medicine–trained clinicians are deployed where their expertise adds the greatest value. Future work should evaluate how different staffing configurations and reimbursement models influence clinician experience, maintenance of clinical skills, clinician retention, and program sustainability. Finally, medicolegal considerations, particularly those related to managing higher-acuity presentations in virtual settings, warrant evaluation as these models evolve.

This study has limitations. The veteran patient population is predominantly older, White, and male, which may limit the transferability of findings to more diverse patient populations. The VA is an integrated national health system which may limit the applicability of aspects of the program to nonintegrated health systems. The national scope of the study sample is a strength, and clinicians who agreed to be interviewed may have been more familiar with TEC, leading to more informed insights. However, although program leads were asked to provide complete staff contact lists rather than nominate clinicians, we could not independently verify the completeness of those lists; eligible clinicians may therefore have been omitted before recruitment. Among those contacted, clinicians who volunteered may also have been more engaged with or favorable toward TEC, introducing additional sampling bias.

Furthermore, our sample of TEC providers limits the transferability of our findings. Our sampling strategy did not capture clinicians who opted out of or discontinued participation in TEC and providers who agreed to participate overrepresented some geographic regions and the static operational model. TEC clinicians reported their perceptions of medical decision-making and job satisfaction; however, this study did not directly measure use, patient safety, or clinician retention outcomes. Finally, since only clinicians were interviewed, perceived improvements in patient-centeredness, satisfaction, and trust do not represent the veteran perspective, and future work should elicit patient perspectives.

Emergency medicine clinicians report that TEC can function to meet veterans where they are, providing evaluation in a virtual setting to prevent avoidable ED visits and addressing access challenges, particularly for rural veterans. For clinicians who combined TEC with in-person emergency practice, this hybrid staffing was perceived as supporting clinician well-being through greater schedule flexibility and more focused time with patients, which participants described as helping them reconnect with their core motivation for practicing medicine. However, this study was not designed to compare staffing models or determine whether one configuration was preferable. Our findings can help inform the development and implementation of similar virtual emergency services and staffing models in other settings.