A descriptive qualitative research design was adopted. The data were collected through semistructured, face-to-face interviews. Given the limited existing knowledge regarding the application and perceptions of VR-based preoperative ICU visits, this design was chosen to capture the complexity of the phenomenon and obtain credible data that closely reflect participants’ genuine thoughts and experiences.

This qualitative study was conducted by a research team from the ICU of a tertiary grade A hospital in Jinan, Shandong Province, China, and was approved by the hospital’s Ethics Committee (approval number 2025‐929). Informed consent was obtained from all participants. All participants were informed that their participation was voluntary and that they could withdraw from the study at any time without any adverse consequences. To protect participants’ privacy, each participant was identified by a code instead of their real name during data analysis. Additionally, no financial compensation was provided to any participant. All original interview data and participant information were anonymized and stored on an encrypted cloud drive to ensure confidentiality.

This study used purposive sampling with maximum variation to ensure the diversity of participant perspectives, which is appropriate for exploratory qualitative research aiming to understand a range of experiences and needs [14]. For health care providers, we intentionally selected participants with varying years of ICU experience (1‐30 y), professional titles (nurse-in-charge to chief nurse, attending to chief physician), and clinical roles (bedside nurses, charge nurses, intensivists) to capture diverse clinical perspectives. For patients’ family members, we recruited individuals across different age groups (26‐70 y), educational levels (primary school to master’s degree), and relationships to patients (spouse, child, parent) to reflect their inherent heterogeneity.

Based on these predefined criteria, potential participants were identified by a researcher with the charge nurses of 2 ICUs at the aforementioned hospital. Eligible individuals were approached in person between April and July 2025, provided with detailed study information, and invited to participate; 3 patients’ family members declined due to time constraints. All selection decisions were reviewed and approved by an ICU nursing expert with 10 years of clinical experience to ensure alignment with study aims.

All participants were fully informed of their rights regarding study participation and provided written informed consent. They were notified that they could withdraw consent at any time without providing a reason or facing any negative consequences. The inclusion criteria were as follows: for patients’ family members being an immediate relative of an ICU patient and fully involved in the patient’s treatment decision-making process; for health care providers: having at least 1 year of working experience in the ICU. The exclusion criterion was the inability of patients’ family members or health care providers to participate in the interview (eg, due to physical or cognitive limitations). To ensure comprehensive reporting, this study adheres to the COREQ (Consolidated Criteria for Reporting Qualitative Research) checklist [15].

Guided by a descriptive qualitative research methodology, the data were collected via in-depth, semistructured interviews. Interviews were conducted face-to-face in either the conference room (for physicians and nurses) or the private conversation room (for patients’ family members) of the ICU. These settings were quiet, comfortable, and private, with no interruptions, to ensure the continuity of thought for both the researchers and participants. The researchers were graduate students who had completed qualitative research courses and possessed interview experience. An initial interview guide was developed following discussions within the research team and consultation with qualitative research experts. A pilot study was then conducted with 1 nurse, 1 physician, and 2 patients’ family members to identify and address potential issues. The final interview guide was finalized after revisions, with details presented in Textboxes 1 and 2. Before each interview, participants were provided with a detailed explanation of the study’s purpose, methods, and confidentiality principles, and they signed the informed consent form on-site. Participants were informed of the concept and purpose of VR-based preoperative ICU visits but did not experience or view an actual VR system or training video. During the interviews, the researchers listened attentively to participants’ responses, promptly asked follow-up questions to clarify ambiguous content, and ensured the completeness and accuracy of the information. Participants were encouraged to share their willingness, needs, and perceptions regarding the VR-based preoperative visit system, with minimal interference from the researchers’ personal views. All interviews were anonymously recorded using a voice recorder, and nonverbal behaviors (eg, facial expressions, body language) of participants were documented simultaneously. Each participant was interviewed once: interviews with patients’ family members lasted approximately 25 to 45 minutes, while those with health care providers lasted approximately 30 to 40 minutes.

Within 24 hours of each interview, the audio recordings were verbatim transcribed into text by the researchers themselves, and the transcribed textual data were returned to the participants for verification, thereby improving the credibility of the data. Subsequently, the transcribed text was imported into NVivo 14.0 software (Lumivero) for analysis, with this process conducted simultaneously and independently by the primary researcher and another master’s student trained in qualitative methods.

This study adopted Braun and Clarke’s 6-phase reflexive thematic analysis framework, which is rooted in a constructivist ontology and interpretive epistemology [16]. This approach posits that themes are actively constructed and interpreted by researchers through reflexive engagement with data, rather than passively discovered as pre-existing entities. The six analytical phases were implemented as follows: (1) familiarization with the data: both researchers independently read each transcript a minimum of 3 times, documenting detailed reflexive notes on initial impressions, recurring meaning patterns, and preliminary insights. This phase aimed to immerse researchers in the dataset and generate initial coding ideas. (2) Generating initial codes: codes were generated inductively directly from the data, unconstrained by preexisting theoretical frameworks or the researchers’ a priori assumptions. Each meaningful text segment (ranging from short phrases to full paragraphs) was assigned a descriptive code capturing its core meaning. A total of 103 initial codes were generated across all transcripts. (3) Searching for themes: similar codes were collated and grouped to construct potential themes, mapping how distinct codes combined to represent broader meaning patterns across the dataset. (4) Reviewing themes: the 2 researchers independently reviewed initial themes to confirm they coherently aligned with the full dataset and that supporting coded extracts were consistent with each theme’s core meaning. Discrepancies were resolved through discussion; a third qualitative research expert was consulted if consensus could not be reached. (5) Defining and naming themes: each theme was clearly delineated and named to capture its core conceptual meaning, with subthemes developed within each main theme to convey nuanced layers of insight. (6) Producing the report: the findings were synthesized into a coherent narrative supported by illustrative participant quotes.

Data collection proceeded separately for each stakeholder group (nurses, physicians, and patients’ family members) given their distinct roles and perspectives. Recruitment and data collection within each group continued until sufficient depth and richness of interpretive meaning were achieved, with no new distinct meaning patterns identified in the final 2 consecutive interviews per group [15]. One additional interview was conducted per group to confirm the sufficiency of the data’s interpretive depth [17].

This study was conducted by a nursing research team with expertise in critical care and qualitative methods. The interviewer was a woman master’s student in nursing with 3 years of ICU clinical experience and formal qualitative research training. None were directly involved in the clinical care of participants’ family members to minimize power imbalances. The corresponding author is an ICU nursing specialist with 10 years of clinical and research experience who supervised the entire study.

Reflexivity was maintained throughout the research process. Interviewers kept reflexive journals to document their preconceptions, and regular team debriefings were held to discuss emerging themes. Independent dual coding by 2 researchers was used to enhance analytical credibility, with discrepancies resolved through consensus discussions.

A total of 29 participants were interviewed between April and July 2025, including 15 ICU nurses, 6 ICU physicians, and 8 patients’ family members admitted to the ICU. Detailed demographic characteristics of the nurses, physicians, and patients’ family members are presented in Tables 1-3, respectively. Four overarching themes were constructed through the reflexive analytical process, reflecting a logical progression from unmet clinical needs to technology adoption barriers for VR-based preoperative ICU visits. Among the 8 patients corresponding to family participants, surgical types included orthopedic trauma surgery, cholecystectomy for gallstones, gastrointestinal resection or anastomosis, partial hepatectomy, and vascular surgery; see Table 4 for details.

This qualitative study directly addresses our 2 primary research objectives: exploring stakeholder willingness and needs for VR-based preoperative ICU visits and identifying core design requirements and implementation barriers. Four overarching themes were constructed: (1) patients’ family members’ perceptions and unmet information needs regarding the ICU, (2) willingness and attitudes toward VR-based preoperative ICU visits, (3) user-centered design requirements for VR systems, and (4) anticipated implementation barriers. All participants held positive attitudes toward VR-based preoperative ICU visits, though their specific needs and priorities differed between health care providers and patients’ family members. Across all groups, 2 core tensions emerged that cut across these individual themes: first, a tension between families’ comprehensive information needs and health care providers’ imperative for workflow efficiency, and second, a tension between participants’ optimistic expectations based on limited VR experience and the practical challenges of real-world use.

This study has several strengths, including the first dual perspective investigation of VR-based preoperative ICU visits in the Chinese context, purposive sampling achieving diversity across stakeholder groups, and rigorous reflexive thematic analysis with independent double coding.

However, several limitations warrant acknowledgment. First, this phase of the study did not include patient perspectives, which represents a significant limitation. This work constitutes the first phase of a multiphase user-centered design project for a VR preoperative visit system. Patients’ family members are core users of the VR system alongside patients, and they also provide essential emotional and informational support to patients during their ICU stay. Meanwhile, health care providers are key stakeholders in implementing any clinical intervention, and their input is critical to ensuring that the system can be integrated into existing clinical workflows. While the perspectives of patients’ family members and health care providers provide essential foundational information for designing a clinically relevant and implementable tool, we acknowledge that the absence of direct patient data limits the completeness of our conceptual framework. Patients are the ultimate end users of the system, and their experiences, preferences, and needs are essential for creating a truly comprehensive patient-centered intervention. Therefore, the design recommendations presented in this paper should be considered preliminary. Patient interviews have already been completed, and the manuscript presenting patient perspectives is currently in preparation. It will be submitted separately to complement the findings of this study, and the insights from the patient study will be used to refine and validate the current conceptual framework, completing the full user-centered design process.

Second, this was a single-center study conducted at a tertiary hospital in eastern China, which limits the generalizability of our findings. Health care systems and cultural contexts vary significantly across regions, and future multicenter studies are needed to confirm our results in different settings.

Third, the uniformly positive attitudes toward VR should be interpreted cautiously. Most participants, particularly patients’ family members, had limited hands-on VR experience, which may have led to overly optimistic perceptions. Additionally, the interview process may have introduced social desirability bias. These findings, therefore, represent preliminary expectations rather than confirmed real-world effectiveness. Limited prior VR experience among participants may have affected their ability to evaluate the technology realistically. As documented in previous research, individuals without hands-on VR experience often have difficulty accurately assessing its practical benefits and challenges. Their perceptions may change significantly once they have the opportunity to use an actual working VR system [25]. Future studies should evaluate the usability and effectiveness of a prototype VR system with actual users.

Fourth, the relatively small number (n=8) of patients’ family members may raise concerns about the comprehensiveness of our findings. However, it is important to note that the rigor of this qualitative study is evaluated by the depth of interpretive meaning and the adequacy of thematic development, rather than by statistical sample size [17]. While sufficient interpretive depth was achieved within this group, a larger sample could provide more diverse perspectives on family caregiver needs.

Fifth, social desirability bias may have influenced participants’ responses. Participants may have been more likely to express positive attitudes toward the technology to please the researchers. We attempted to minimize this bias by emphasizing that there were no right or wrong answers and that all opinions were valuable.

Sixth, this study relied solely on semistructured interviews and did not include methodological triangulation with observations or usability testing, which could have enhanced the validity of our findings.

Despite these limitations, this study provides critical insights into stakeholder perspectives on VR-based preoperative ICU visits, laying a solid foundation for system development and supporting further evaluation of its real-world feasibility and effectiveness.

This study identifies a critical gap in perioperative care: patients’ family members receive detailed information about surgical risks but almost no formal education about the ICU environment or postoperative care processes. Across all stakeholder groups, participants expressed positive anticipated acceptance of VR-based preoperative visits, recognizing their potential to address this gap through immersive, standardized information delivery.

These findings suggest that effective VR design must reconcile 2 potentially competing priorities: comprehensive environmental and procedural simulation that meets families’ information needs, and seamless integration into existing ICU workflows without adding to staff burden. The former underpins acceptance; the latter determines sustainability.

The immediate next step is prototype development followed by usability testing with actual end users, as the optimistic expectations expressed by participants with limited prior VR experience may not persist with hands-on use. If validated, VR-based preoperative ICU visits could offer a scalable solution to a persistent communication challenge in perioperative care.