This was an open-label randomized controlled trial conducted between March 2022 and January 2023 at Radboud University Medical Center, Nijmegen, the Netherlands. The trial is reported according to the CONSORT (Consolidated Standards of Reporting Trials) guidelines for randomized trials (Multimedia Appendix 1) [24].

Ethical approval was obtained by the Research Ethics Committee of the Radboud University Medical Centre, the Netherlands (Commissie Mensgebonden Onderzoek Arnhem-Nijmegen, NL80011.091.21). Eligible patients were approached by their treating physician or nurse and received verbal and written information about the study. A member of the research team answered any additional questions and obtained informed consent. All participants received a unique research ID linked to their personal details. The participants’ data were securely stored on the institution’s server, and access to the specific folders was restricted to the researchers involved in the study. The study data were stored using only the research ID, with no personal identifiers, in Castor Electronic Data Capture (Amsterdam, the Netherlands). Participants received no reimbursement. This trial was conducted according to the principles of the Helsinki Declaration and in accordance with Dutch guidelines, regulations, and Acts (Medical Research involving Human Subjects Act, WMO).

The study population comprised of patients aged 16 years and older, with no upper age limit, who had a blunt chest trauma and were directly admitted to the trauma and orthopedic ward from the emergency department. The study included all genders to account for potential variations in pulmonary mechanics. Exclusion criteria were (1) neurotrauma with Glasgow Coma Scale of 13 and lower; (2) history of dementia, seizures, and epilepsy; (3) significant hearing or visual impairment which is not corrected; (4) headwounds or damaged skin with which comfortable and hygienic wear of head-mounted display (HMD) is not possible; (5) stay at ICU during current hospital admission for reasons other than observation or a duration of 48 hours or longer; and (6) erect position in bed is not possible or allowed. Patients were recruited on the first day of their admission. Patients were subsequently randomized (1:1) to the VR group or control group using computer-generated block randomization.

All patients from both groups received care according to the existing guideline [25]. Similarly, pain management was according to the World Health Organization guidelines for acute pain in adults [26]. Once daily, the control group received usual physiotherapy consisting of breathing and physical exercises. Breathing exercises include deep breathing, huffing, and coughing following the active cycle of breathing technique. Patients were instructed by the physiotherapist and received a leaflet with written instructions. Physical exercises include practice of functional movement (eg, activities of daily living) and exercises for range of motion of the trunk and limbs. Patients were instructed to perform breathing exercises 8 times daily for 10 minutes and to extend these exercises 2 times daily with (sitting) physical exercises for an additional 10 minutes.

The VR group was instructed to perform the breathing exercises using the VR intervention 8 times daily for 10 minutes and to extend these exercises 2 times daily with (sitting) physical exercises using VR for an additional 10 minutes. For all VR exercises an HMD, the PICO Neo 3 (Pico) was used. The commercially available apps SyncVR Fit and SyncVR Relax & Distract (SyncVR Medical) were used. This selection of apps was made in consultation with physiotherapists and nurses to ensure that the games were suitable and beneficial for the patients. This collaborative approach aimed to choose games that would effectively address the therapeutic objectives of pain relief, adequate breathing, and rapid mobilization. SyncVR Fit contains breathing exercises that are comparable to the exercises in usual physiotherapy care but can be performed in a virtual environment. The physical VR exercises consist of several games through which patients are challenged to reach out to objects while involving their arms, head, and trunk. Once daily, the exercises were performed under the supervision of a physiotherapist, the other sessions were unsupervised. SyncVR Relax & Distract contains several mindfulness and relaxing exercises to support in coping with pain and anxiety. Patients were allowed to use these at their own discretion in addition to the prescribed exercises with a maximum of 30 minutes per VR session to prevent side effects.

In both groups, a research team member visited the patients twice daily to perform measurements. At these moments, patients were encouraged to perform the exercises in attendance of the research team member. In the VR group, this was primarily to detect any technological difficulties and lower the bar for patients to use the VR headset. The control group was visited accordingly to prevent bias.

Primary outcome was the vital lung capacity (in mL), measured using an incentive spirometer (IS; Voldyne [Teleflex Medical]), on day 5 of the study or the last measurement between 3 and 5 days. Incentive spirometry was chosen over usual spirometry despite lower reliability because transferring patients to a spirometer at least once daily would have required significant logistical support and funding. In addition, it would have imposed an additional burden on both patients, who were sometimes quite ill, and the already busy personnel. To ensure consistency in the measurements, a standardized protocol was followed, which included recording the best score out of 3 attempts. Secondary outcomes were patient mobility, analgesics use, clinical outcomes, pain, activities of daily living, patient-reported outcome measures, safety outcomes, and barriers and facilitators toward implementation. Patient mobility was defined as the percentage of time spent lying, sitting, and moving measured using a wearable activity monitor (activPAL3 [Pal Technologies Ltd]). The type and dosage of analgesics used were extracted from patient files for the first 5 days of the study. Opioid use was calculated and reported as oral morphine equivalent. Clinical outcomes included the length of hospital stay (in days), pulmonary complications during admission, transfer to ICU, and readmission within 30 days. Pain was assessed during breathing exercises using a visual analog scale (VAS) with 0 being no pain and 10 being extreme pain. Activities of daily living were measured using the powerlessness in daily living (PDL) questionnaire, with lower scores representing more independence in activities of daily living [27]. Patient-reported outcomes and experiences comprised the Quality of Recovery-15 (QoR-15) questionnaire [28], modified treatment satisfaction questionnaire (MTSQ) [29], and semistructured interviews regarding patients’ satisfaction and experiences with VR. Reasons for withdrawal and side effects during VRx were registered as safety outcomes. Barriers and facilitators were derived from the semistructured interviews and patient diaries in which patients reported on treatment adherence, technical problems, and feedback on the VR exercises.

The duration of the study for patients was 5 days from the moment of inclusion on the first day of their admission or less when the patient was discharged earlier. After randomization, participants in the intervention group received instructions from a research team member on how to use the VR intervention, and a supervised training session was performed. If needed, additional technical support was provided by the research team. Patient characteristics were extracted from electronic patient files and additional characteristics were asked upon inclusion. A wearable activity monitor was installed on the first day. At baseline IS, VAS pain score, and PDL score were measured. During days 2-5, daily IS, VAS pain score during breathing exercises, PDL score, and patient mobility were registered for both groups. IS was measured twice daily following a standardized procedure. Patients were instructed to register the frequency of pulmonary and physical exercises in a daily diary as well as to fill out the quality of recovery questionnaire and any experienced side effects (open-ended question). On day 5, participants received the treatment satisfaction questionnaire. Patients in the intervention groups that used the VR intervention at least once were asked on day 5 to participate in a short semistructured interview with the predetermined topics, such as experiences with VR regarding efficacy and barriers and facilitators using VR from patients’ perspectives. After 5 days, participants in the VR group were allowed to choose to continue the VR intervention or to return to usual care. Patients were followed up for 30 days after discharge to register clinical outcomes such as late complications. In Multimedia Appendix 2, all study procedures are schematically presented.

The calculated needed sample size was 63 patients per group. This calculation was based on the primary outcomes measure vital lung capacity. The mean outcome of the control group was set on 1250 cc and SD was set on 500 cc based on the literature [30]. The mean outcome of the VR group was based on an expected 20% (clinically relevant) improvement, resulting in 1500 cc. An α of .05 and β of .20 were applied.

Quantitative analyses were done using IBM SPSS Statistics (version 25). Qualitative analyses were done using Atlas.ti (version 23.0.7; Lumivero). Descriptive statistics were used to present quantitative data. For effectivity outcomes, only patients who completed at least 3 study days were eligible for evaluation. No significance was calculated due to the low number of participants. For safety outcomes (eg, side effects), all patients were evaluated. Mobility data derived from the activPAL devices are automatically checked for validity by built-in algorithms following the validation criteria of PAL technologies [31]. This validity is indicated per measurement day, meaning that all measurements for a given day are regarded as either valid or invalid. This applies uniformly to all mobility measures, including standing, lying or sitting, and walking. Only valid wear days were evaluated.

The content of the interviews was transcribed ad verbatim and analyzed in Atlas.ti (version 23) using inductive thematic analysis. To minimize bias, 2 independent researchers with a background in medicine performed thematic analysis. Transcripts were coded according to the predefined main topics. Identified codes were categorized into themes and discussed until a consensus was reached.

Between March 2022 and February 2023, a total of 129 patients were screened for eligibility and 27 patients were recruited (Figure 1). Reasons for exclusion are described in Table 1. In the VR group, 2 patients did not receive the intervention; 1 patient withdrew before the intervention was started, and 1 patient was excluded on day 2 due to previously unrecognized cognitive impairments. In the VR group, 5 patients did not complete 3 days; 1 patient stopped early because he found the exercises too time demanding, 1 patient stopped because she was unable to follow study protocol, and 3 patients in the VR group were nonevaluable due to discharge before completing 3 study days. In total, 19 patients were evaluable (completed 3 or more days) of which 7 in the intervention and 12 in the control group. In total, 4 patients in the intervention group and 8 patients in the control group completed the full 5 days of the study protocol. Majority of evaluable patients were male (15/19, 79%). The participants’ mean age was 60 years. Mean Injury Severity Score was 15 (SD 7; Table 2). Adherence to exercises was poor in both groups, with an average of performing breathing exercises 2 times per day in the VR group and 3 times per day in the control group (Table 3).

Vital lung capacity at 5 days (or last measurement between day 3 and 5) was similar between groups with 1830 (SD 591) mL in the control group and 1857 (SD 435) mL in the VR group. The vital lung capacity over the days improved with an average of 421 (SD 345) mL between day 1 and 3 in the VR group followed by a decrease of 208 (SD 315) mL between day 3 and 5, which is likely due to early discharge of rapidly recovered patients (Multimedia Appendix 3). In the control group, an average improvement of 389 (SD 370) mL was observed between day 1 and 3 with a further increase of 246 (SD 292) mL between days 3 and 5.

Regarding the mobility data measured using activPal, 58% (24/37; SD 36) of measurement days was valid in the control group and 61% (13/21; SD 34) was valid in the VR group. Mobility was similar between groups (Multimedia Appendix 3) with majority of time spent sitting (in bed) or lying. Patients were standing or walking for only 3%-5% of the time in both groups on all days. Analgesics use was similar between groups (Multimedia Appendix 4). In both groups, there was a slight improvement in the VAS pain score during breathing exercises over the course of 3 to 5 days (Multimedia Appendix 3). For the PDL questionnaire, a similar pattern was found as for vital lung capacity with initial improved median scores in the VR group from 9 (IQR 8-12) points to 5.0 (IQR 4-5) points between day 1 and 3 and from 11 (IQR 4-15) to 4 (IQR 0-13) in the control group within both groups worsening between day 3 and 5 (Multimedia Appendix 3). Regarding the quality of recovery, the control group scored median daily scores between 83.0 (IQR 76-97) and 99 (IQR 89-111) compared with 81 (IQR 57-111) and 108 (IQR 67-110) in the VR group (Multimedia Appendix 3). Mean global satisfaction was 72.9 (SD 12.5) points on a scale from 0 to 100 in the VR group versus 65.6 (SD 15.7) in the control group. Median length of hospital stay was 5 days in the VR group and 6 days in the control group. No pulmonary complications, ICU admissions, or readmission within 30 days were observed in both groups.

A total of 6 patients of the VR group were interviewed—4 males and 2 females. Mean age was 57 years. Interviews were conducted face-to-face or by telephone. The interviews had a median duration of 11.5 (range 5.5-14.5) minutes. In the interviews, 5 themes were identified within the 2 main topics (Table 4).

This study was ended prematurely due to an insufficient rate of accrual and operational futilities. At start of the study, the tail of the COVID-19 pandemic caused a delay in inclusion rate due to reduced bed capacity. After the first 3 months, the number of screened patients caught up with the planned screening numbers; however, the rate of accrual remained insufficient. Based on the actual recruitment rate, 5 additional years of recruitment would have been required to attain the necessary number of participants (n=126). This urged a critical appraisal of the inclusion procedure and consideration to expand the target patient population to other wards. Simultaneously, it was noticed during study conduct that adherence to the prescribed frequency of breathing and physical exercises was low in both the VR and control groups. This would result in a smaller effect of VRx than expected and a smaller difference with the control group. Consequently, the study would require a larger sample size to obtain a significant difference, while it would be questionable whether this difference would be clinically relevant. The low adherence in the control group prompted us to audit the protocol adherence at several wards with a comparable breathing exercise protocol. Altogether, we drew the conclusion that it would require an additional intervention to optimize protocol adherence, which was considered unrealistic and inappropriate in the current study.

The primary aim of this study was to assess the effectivity of VR breathing and physical exercises on the pulmonary recovery of patients with blunt chest trauma at the trauma ward. In this small sample, we observed no differences in final vital lung capacity between the VR group and control group. The pattern of vital lung capacity increase seemed to differ between groups, with the VR group reaching a higher capacity on day 3 than the control group. A similar pattern was observed for the independence in activities of daily living. Patients reported outcomes on quality of recovery and satisfaction did not differ between groups. Interviews demonstrated appreciation and potential of VR exercises, although several barriers were mentioned regarding feasibility and usefulness.

Despite a preceding audit of this patient group regarding eligibility and sufficient numbers, the trial was prematurely terminated due to enrollment failure. Enrollment failure is the main reason for termination in 60% of prematurely terminated trials [32,33]. A secondary reason for termination was poor adherence to the clinical guideline in both the control group and VR group with infrequent to no performance of breathing and physical exercises in both groups. We considered this a major drawback for continuation with this study protocol considering mostly unsupervised and self-administered exercises in both groups.

No conclusions can be drawn about the effectivity of VR exercises on pulmonary recovery of patients with blunt chest trauma due to the small obtained sample size. However, the results can be interpreted as those of a pilot randomized controlled trial, and several important lessons can be learned for future VR studies in similar and different contexts. First, the suitability of patients to use VR in a hospital (trauma) ward setting might be lower than generally expected. Before initiation of this trial, calculations were performed based on the hospital registry of the year 2020 of patients who were admitted to the hospital with blunt chest trauma for more than 24 hours and no ICU admission. We estimated that 60% (146/243) of these patients could be included based on a previous clinical study in our hospital [34]. However, only 21% (27/129) of these patients could be included in this trial. Main exclusion reasons were a headwound, inability to comprehend the study protocol (eg, language barrier), cognitive impairment, and delirium, together accounting for 44% (45/102) of patients excluded. Another 24% (24/102) declined to participate for various reasons such as feeling too sick, disinterested, and finding participation too burdensome. Especially, the proportion of patients excluded due reasons related to inability to use a VR headset was underestimated in designing this study. We did not investigate impact of age as barrier or facilitator for using VRx. However, contrary to what has been reported in literature, we did not observe any noticeable resistance or lack of motivation among older patients during the inclusion procedure or the intervention [35]. In literature, it is rarely described how many patients are eligible for VR use and the reasons for exclusion, such as inability to use a VR headset due to concomitant illness. For example, 10,776 patients were assessed for eligibility in a study of Wiechman et al [36] evaluating the effectiveness of VR for pain and anxiety management in trauma patients and only 184 (2%) were included. Spiegel et al [17] showed an inclusion rate of 23% in a general hospital population. In both studies, specific reasons for exclusion are not mentioned and are generically described as “not meeting inclusion criteria” and “declined to participate.” Future research should specifically report on the eligibility of patients and exclusion because of the inability to use VR.

Second, the setting in which VRx is applied should be thoroughly charted before conducting research or implementing VR. Although a physiotherapist and nurse were involved in the design of this study to ensure alignment with daily practice, adherence to protocolled exercises was unexpectedly low. A study of Martin et al [37] in postoperative patients showed that 26% of patients failed to use IS correctly and 38% denied ever using IS. This is consistent with our study in which on different days 33% (4/12) to 63% (5/8) of the patients in the control group used IS 2 times or less, which were likely the supervised exercises for measuring vital lung capacity. Martin et al [37] reported that following a brief educational intervention by a physician, 74% of patients were more confident to use IS during the remainder of their care. In our study, patients were encouraged and educated once daily by a physiotherapist to perform the breathing exercises, however adherence was still low. Literature shows that self-management and patient participation in the hospital can improve treatment adherence [38]. The intervention in this study was meant to engage patients in their treatment, however this did not result in improved treatment adherence. Generally, VR is considered appropriate as self-management tool for patients [39]. In this study, patients acknowledged this potential, but the intervention did not result in improved treatment adherence. It has been argued that VRx still requires professional guidance [40,41]. This professional guidance might even be of greater importance in a hospital environment, since patients have trouble using VRx independently due to illness and comorbidity. Nurses may play an important role in giving education, counseling, facilitating, and enhancing taking responsibility [42]. However, a high workload and contradicting patient expectations are factors that complicate patient engagement [43,44]. The low adherence in our study might imply that the hospital environment, including patients and caregivers, needs reorganization to allow the transferring of care possibilities to patients and supporting patients in acquiring the self-management skills needed for VRx [45]. Better education and training of both caregivers and patients could enhance successful implementation of interventions such as VR [40,45,46]. We cannot rule out that results are different for a setting with close monitoring by caregivers, for example, high care unit or highly disciplined patient groups such as injured military personnel.

Third, several identified barriers should be overcome to ensure successful deployment of VR on the hospital ward. The barriers were identified by interviewing patients focused on their user experiences. Over the last decade, collecting patient experiences has been emphasized as a starting point for improving patient care in general [47]. This study illustrates how experiences can reshape new innovations like VRx for blunt chest trauma. Patients mentioned the perceived usefulness relative to the system usability as important reason for low adherence and as a main barrier to self-managing the HMD and the different apps. As described in the Technology Acceptance Model designed by Sagnier et al [48], the intention to use a given technology is predicted by the perceived usefulness and the perceived ease of use. It was clear from reported experiences that the perceived usefulness did not outweigh the perceived ease of use in the interview group. Although many feasibility studies report on the acceptability of VRx, a systematic analysis of the acceptance of VR is rarely performed [48]. Furthermore, in feasibility studies, acceptability is defined by a variety of different outcome variables such as a sense of involvement, comfort, wish to use VR again in the future, withdrawal from study, and satisfaction [49-52]. A systematic analysis, for example, using the Technology Acceptance Model, may add to the knowledge derived from pilot and feasibility studies in medical VR. Another barrier was the busy hospital ward, albeit that some purposively took on the VR device to escape from the busy environment and pursue quietness and privacy. A similar finding was reported for VR postoperative pain management [53]. The results underline the different values of patients regarding the use of a digital technology and prompt the alignment of outcome measures regarded as relevant in the design and evaluation of a VR intervention [54,55].

The relevance of this study lies in underlying causes for early termination, the critical appraisal of the study setting and standard treatment in the control group, and the yield of the patient reported experiences from the interviews. Main lesson was the misassumption that patients adhere to the instructions in the hospital protocol for breathing and physical exercises and that physiotherapists, nurses, and physicians check protocol compliance. A preceding pilot study might have exposed inadequate compliance as well as the overestimation of eligible patients. However, we doubt such a pilot study would have revealed the barriers regarding integration of VR in the clinical workflow and the controversies in perceived usefulness between patients.

Several recommendations can be made for research and implementation of clinical VR from the appraisal of our findings. First, adherence to study protocols should be charted before designing a VR clinical trial. This may be accomplished by “shadowing” clinician-patient interaction and patient’s functioning and well-being [56]. Simultaneously, data can be obtained for integration of VR treatment in the daily workflow and additional training of staff [40,46]. Second, individual patient reported experiences and values need to be prioritized in evaluating VR acceptance, usability, and effectiveness [53]. In line, we recommend to perform a systematic analysis on the acceptance before pilot or main effectiveness studies [48]. Third, eligibility of patients and exclusion of patients due to the inability to use VRx should be routinely reported.

This clinical trial of self-administered VR treatment for blunt chest trauma had to be terminated prematurely due to enrollment failure and limited protocol compliance to breathing and physical exercises in both groups. Suitability of trauma patients to use VRx at a hospital ward was overestimated despite previous audit of potentially eligible study participants. Hospital setting, standard care, and patients’ perceptions of VR treatment seem important determinants for success in clinical VR research.