Pain Rehabilitation Virtual Reality (PR-VR) at Home: A Pilot Randomized Controlled Trial for Youth with Chronic Pain

ABSTRACT

Background

Pain Rehabilitation Virtual Reality (PR-VR) is a novel tool designed to engage youth with chronic pain in exercise as a part of physiotherapy. PR-VR has been evaluated in day-treatment and outpatient settings, but not for at home use.

Aims

To evaluate the feasibility, acceptability and safety of PR-VR at home for youth with chronic pain. An exploratory aim evaluated the preliminary effectiveness of PR-VR on pain and disability.

Methods

This was a 3-arm pilot randomized controlled trial (registration #NCT04912817). Each study arm included four 60-min physiotherapy sessions via Zoom (1: commercially available VR; 2: custom-developed PR-VR; 3: standard physiotherapy care). Feasibility metrics included enrollment, accrual, and drop-out rates. Acceptability was measured via satisfaction surveys and qualitative interviews. Pain and disability were measured via self-report, physical performance, and actigraphy.

Results

Thirty-seven of 70 eligible patients approached were enrolled and randomized (53% enrollment rate, 100% accrual rate), including 28 females (76%) aged 10–17 years. Twenty-eight participants completed the study (33% drop-out rate; reasons include pain resolution, medical complexity, dissatisfaction with assignment to control). Ten minor adverse events were reported across study arms. Participants and physiotherapists expressed high levels of VR acceptability. Of the participants who completed arms 1 and 2 (n = 17), 94% were satisfied/very satisfied with VR and 100% rated VR as easy/very easy to use.

Conclusions

Results from this study demonstrate the safety and acceptability of VR in pediatric pain rehabilitation. These feasibility results will inform the design of future large-scale trials, including consideration of a wait-list control group.

Trial Registration

ClinicalTrials.gov, NCT04912817

Introduction

Chronic pain affects 1 in 5 Canadian children and youth, posing a significant burden to the economy and individuals living with pain.¹ The prevalence of pediatric chronic pain peaks during adolescence, a period when pain and disability can interfere with education and employment,² and when persistent into adulthood, is linked with long-term disability, reduced work participation, and increased health care utilization.^3,4 Effective management of pediatric chronic pain is critical to potentially modify a trajectory of pain and disability.

Gold standard treatment for pediatric chronic pain is a multi-modal interdisciplinary approach combining pharmacological, psychological therapies and physical therapies (PT).² These treatments focus primarily on restoring a child’s function and minimizing disability, in addition to pain reduction. In contrast to PT for acute pain (e.g., heat, ice), with chronic pain, active PT strategies (e.g., exercise, movement) are emphasized.^2,5 Engaging in exercise can be particularly challenging for youth with chronic pain who fear movement due to risk of increased pain or reinjury.⁶ Fear of movement can lead to activity avoidance, deconditioning, and further pain, reinforcing a cycle of disability. To address fear of movement, effective exercise interventions for youth with chronic pain are ideally age-appropriate and play-based.⁵

Virtual reality (VR) technology offers an innovative approach to engage youth with chronic pain in exercise in a way that is fun, distracting and interactive.⁷ Using head-mounted displays and handheld controllers, the user is transported to a virtual environment that simulates real-world tasks or playful challenges.⁸ Commercially available VR software (e.g., Beat Saber) that incorporates exercise designed for entertainment or fitness can be used in PT to target movements such as squatting, reaching, or upper limb movement. However, applying commercially available software in PT for youth with chronic pain is limited in that the exercise parameters may be too difficult or the lack specificity to target movements of the painful body part(s).

While custom VR applications have been widely used and tested to treat pediatric acute pain, with promising results to reduce fear and pain intensity, few studies have examined VR to treat pediatric chronic pain.⁹ To date, Pain Rehabilitation Virtual Reality (PR-VR) is the only custom VR program created specifically to enhance PT for pediatric chronic pain.¹⁰ Implementation of PR-VR as part of face-to-face PT in a pediatric pain day treatment setting was found to be acceptable, safe, highly engaging, and was associated with increased patient movement and decreased pain, fear, avoidance, and functional limitations.¹⁰ However, implementation with outpatient PT faced additional barriers including system-level structural barriers regarding billing and compensation, limited physiotherapist training of VR technology, and participant expectation violation.¹¹

Access to pediatric pain clinics in Canada is limited, with long wait times and regional gaps in service.^12–14 At the onset of the COVID-19 pandemic, many pediatric pain clinics pivoted to virtual care delivery. Although public health pandemic restrictions have since lifted, virtual care remains popular with its ability to mitigate geographic barriers to care.¹⁵ However, delivering virtual PT for youth with chronic pain is challenging given the exercise-based approach to treatment with requisite equipment needs.

To address this challenge, we adapted PR-VR for at-home use, allowing this technology to be used as part of virtual PT in an outpatient pediatric pain clinic. The aim of this study was to evaluate the feasibility, acceptability, and safety of PR-VR for youth (8–18 years) with chronic pain in a 3-arm pilot randomized controlled trial (RCT) comparing commercially available VR, customized PR-VR, and standard PT. An exploratory aim was to evaluate the preliminary effectiveness of PR-VR on pain and disability.

Materials and methods

Study design and setting

This study was designed as a three-arm pilot RCT with 1:1:1 allocation comparing commercially available VR (off-the-shelf VR), customized PR-VR software (custom VR) to a control group (standard PT) as part of outpatient PT for youth with chronic pain. See Fig. 1 for study design and randomization schema. The trial protocol was registered (#NCT04912817) on Clinical Trials (weblink: clinicaltrials.gov/study/NCT04912817?term = jennifer%20stinson&rank = 5) prior to commencement of the study (May 31, 2025).

Figure 1.

Study design and randomization schema.

NRS = Numerical Rating Scale; PROMIS = Patient-Reported Outcome Measurement Information System; PCS-C = Pain Catastrophizing Scale Child; FOPQ = Fear of Pain Questionnaire; TSK-6 = Tampa Scale Kinesiophobia 6-item short form; LEFS = Lower Extremity Functional Scale; UEFI = Upper Extremity Functional Index; STS = Sit to stand; 6MWT = 6 Minute Walk Test; SLB = Single leg balance; CKCUE = Closed Kinetic Chain Upper Extremity; RPE = Rating of Perceived Exertion; PRPS = Pittsburgh Rehabilitation Participation Scale; CPM = Child Presence Measure.

The study was conducted at the Chronic Pain Clinic at the Hospital for Sick Children (SickKids) in Toronto, Ontario, a tertiary-care outpatient pediatric pain clinic that provides care to youth with chronic primary pain conditions. The study was approved by the SickKids Research Ethics Board (1000080939). This trial was designed in accordance with the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) guidelines and the manuscript adheres to the Consolidated Standards of Reporting Trials (CONSORT) reporting guidelines for randomized trials (Supplemental File 1).^16,17

Participants

Patients were considered eligible if they were English-speaking youth (8–18 years) with chronic pain on a stable medication protocol who required outpatient PT for pain management. All participants were approved to participate in the study by their clinic physician and were deemed safe to engage in exercise under supervision with a physiotherapist. Participants required an internet-enabled device to engage in virtual PT and must live within 150 km from the hospital (to easily obtain and return study hardware). Patients were excluded if they had a major untreated psychiatric illness or medical comorbidity that precluded the use of VR (e.g., seizure disorder, taking sedating medication, new onset migraines). For a comprehensive list of inclusion and exclusion criteria, see Supplemental File 2.

Study recruitment and allocation procedures

Potential participants were identified by a Research Assistant (RA) by screening weekly clinic lists for eligibility criteria. Participant eligibility was confirmed by a member of the patients’ care team. All participants provided written, informed consent electronically via REDCap.¹⁸ Consenting participants completed baseline (T0) measures (self-report and performance-based measures with a trained outcome assessor) and were then randomized to one of three study arms: Arm 1: off-the-shelf VR, Arm 2: custom VR, or Arm 3: control. Randomization was computer-generated via the REDCap randomization module with concealed allocation. Outcome assessors and data collectors were blinded to treatment assignments. The RA who enrolled participants did not have access to the random allocation sequence and was a separate individual from the individual who allocated participants to each study arm.

Each participant assigned to a VR study arm participated in an orientation session with an RA to review how to use the VR hardware (e.g., WIFI set-up, login) and ensure safe use (e.g., clearing the space of any obstacles).

Interventions

Across all study arms, participants completed four virtual PT sessions, each 60 minutes in length, approximately every 1–2 weeks. Sessions were delivered by a physiotherapist and/or physiotherapist assistant. The intervention duration and frequency were chosen to align with the physiotherapist’s clinical practice in the recruiting clinic. At the start of each session, the physiotherapist performed a safety check to confirm adequate space for exercise and asked screening questions to identify significant new health concerns or medication changes. All virtual sessions occurred via Zoom for Healthcare. Participants and physiotherapists were not blinded to the intervention arm given the active nature of exercise interventions.

Arms 1 and 2 – VR: For participants in the VR study arms (off-the-shelf and custom VR), PT sessions included approximately 30 minutes of VR (including set-up, tear-down, and rest breaks if needed) and 30 minutes of standard care. Oculus Quest 2 VR hardware was used. Participants were instructed to remove the headset and stop immediately if they experienced dizziness, nausea, or headache and to notify the physiotherapist. Participants were instructed to use the VR device only when a caregiver (≥18 years) was present in the home to ensure safe use. Caregivers did not otherwise participate in study activities. Participants were prescribed a home exercise program, including VR, for daily use. Standard care included review of pain and symptoms, review of treatment goals and home exercise programs as well as adjunctive approaches such as mirror therapy or desensitization, as indicated.

Arm 1 – Off-the-shelf VR: Participants used commercially available games designed for physical activity, however not specifically for PT. Games included Beat Saber, Tilt Brush, Fruit Ninja, Pro Putt, and Dance Central. These games target movements such as squatting, stepping, and upper extremity exercise, cardiovascular exercise, and were tailored by the physiotherapist to the participant’s needs. As such, while these games can be adapted for clinical use by physiotherapists, they were originally designed for entertainment rather than therapeutic purposes. Physiotherapists selected games based that best aligned with the patient’s goals (e.g., Beat Saber engages youth in cardiovascular exercise, Fruit Ninja focuses on upper extremity gross motor movement) and interest.

Arm 2 – Custom VR: Participants used custom PR-VR software developed by the Stanford Chariot Program (https://chariot.stanford.edu/) in collaboration with pediatric rehabilitation providers with the intention for use in PT.¹⁰ PR-VR software “Fruity Feet” included add-on games Alien Invaders and Space Burgers 2. These were designed to promote movement and exercise through playful, interactive tasks. This software is customizable to the participant in that the physiotherapist or user can modify the parameters to target the painful area (upper extremity, lower extremity, left, right) and adjust the difficulty level. Space Burgers incorporates leg ergometer cycling (DeskCycle 2). A detailed description of the software and customizable settings is described elsewhere.¹⁰ The software was initially developed for in-hospital use, using HTC Vive Hardware, which necessitated a wired connection from the head-mounted display to a laptop, with motion sensors mounted to ceiling walls. For this study, we adapted the PR-VR software for compatibility with Oculus Quest, allowing for a more economical application for at home use. For participants in this study arm, commercial VR software (described above) was also available for participants’ recreational use.

Arm 3 – Control: Standard physiotherapy treatment included goal setting, graded exercise, education, prescription of a home exercise program, as well as adjunctive approaches such as mirror therapy or desensitization, as indicated.

Study measures

The study schema is presented in Fig. 1. Participants completed electronic self-report study measures at baseline (T0), immediately following each PT session (T1), and end of study (T2). Physiotherapists completed electronic study measures after each session, (T1). T1 consisted of up to four PT sessions (PT1, PT2, PT3, PT4), which are hereby referred to as T1. Participants completed performance-based study measures at T0 and T2 with a trained outcome assessor. Actigraphy was measured throughout the intervention period (T1-T2).

Participant characteristics: At T0, participants reported their demographic characteristics (age, sex, race/ethnicity).

Feasibility: The RA tracked feasibility metrics including enrollment rate (number of participants enrolled/approached), attrition rate (number of participants who dropped out/enrolled), completion rate (number of participants who completed endpoint measures/enrolled), and adherence to study intervention (# study visits completed). Reasons for study drop-out and reasons for visit cancellation/no-show were logged by the study RA.

Acceptability: Participants completed satisfaction surveys throughout the study period. At T0, participants were asked about their treatment expectancy treatment (e.g., “how successful do you think the treatment will be in improving your functioning?,” “does the therapy offered to you make sense for treating your pain?” rated on a 0–10 scale [0 = not at all, 10 = very much]). At the end of every session (T1), participants rated their ongoing satisfaction with the intervention (e.g., “I enjoyed my physiotherapy exercises,” “I forgot everything around me during my physiotherapy exercises” rated on a 0–100 scale [0 = strongly disagree to 100 = strongly agree]). At this timepoint (T1), participants also rated their level of engagement including ratings of exercise intensity (patient-rated Borg Rating of Perceived Exertion¹⁹) and VR immersion and presence (patient-rated Child Presence Measure²⁰, administered only to participants in the VR study arms). Physiotherapists also rated participant engagement using the Pittsburgh Rehabilitation Participant Scale.²¹). At endpoint (T2), participants rated their overall satisfaction with the intervention (e.g., how satisfied were you with the services you received during the study?), and participants in the VR groups completed an additional survey regarding their satisfaction with VR (e.g., “how easy was it to use the VR device and applications?”).

Participants in the VR study arms were invited to engage in a semi-structured interview to further explore their experience with the intervention. The study physiotherapists engaged in a focus group when the study was completed to explore their satisfaction with the study protocol and interventions. See Supplemental File 3 for a copy of the patient and physiotherapist interview guides.

Safety: Physiotherapists were asked to report adverse events during each virtual PT session in a survey by selecting from a list of known events (including eye strain, dizziness, lightheadedness, distress, nausea, disorientation, seizure) as well as open-ended questions. Participants were asked to report adverse events outside of PT sessions by directly contacting the RA.

Preliminary effectiveness: Measures to estimate preliminary effectiveness included self-report and performance-based measures, measured at T0 and T2. Self-report measures included measures of pain intensity (Numerical Pain Rating Scale [NPRS]²²), pain interference (PROMIS Pediatric Pain Interference²³), pain catastrophizing (Pain Catastrophizing Scale – Child [PCS-C]²⁴), fear of pain (Fear of Pain Questionnaire²⁵), fear of movement (Tampa Scale of Kinesiophobia – 6 item [TSK]²⁶), activity limitations (Lower Extremity Functional Scale [LEFS],²⁷ Upper Extremity Functional Index [UEFI],²⁸ PROMIS Pediatric Mobility²⁹), and Patient Global Impression of Change (PGIC).³⁰

Performance-based measures of physical function included Sit-To-Stand (# reps in 30 seconds),³¹ Single Leg Stance (# seconds with eyes open and closed), Six-minute Walk Test,^32,33 and Closed Kinetic Chain Upper Extremity Test.³⁴ The outcome assessor was a RA trained by a physiotherapist to administer the performance-based measures; this RA was blinded to the intervention arm which participants were assigned. Real-time monitoring of physical activity was measured using Actigraphy. Participants wore watches from T1 to T2 and were instructed to continuously wear them except during water-related activities. Study measures are listed and described in Table 1, with more detailed descriptions of reliability and validity available in Supplemental File 4.

Table 1.

Study measures.

Measure	Description
Self-report
Lower extremity functional scale (LEFS)27	20-item self-report tool that measures lower extremity activity limitations (e.g., squatting, walking, running). Total score ranges from 0–80, where higher scores indicate fewer activity limitations. SEM = 5 points, MDC = 9 points; MCID = 9 points.
Upper Extremity Functional Index (UEFI)28	20-item self-report tool that measures upper extremity activity limitations (e.g., lifting, opening doors) Total score ranges from 0–80, where higher scores indicate fewer activity limitations. SEM = 4.0; MDC = 9.4; MCID = 8.
Fear of Pain Questionnaire Child25,37	10-item tool that measures fear of pain. Total scores range 0–40, with higher scores indicating greater pain-related fear and avoidance.
PROMIS Pediatric Mobility – SF (8a)29	8-item tool that measures impairments in mobility (e.g., climbing stairs, standing). Raw scores are converted to T-scores (range 0–100, mean = 50, SD = 10). Lower scores indicate greater impairment in mobility. MCID = 3 T-score points.
PROMIS Pain Interference – SF (8a)29	8-item tool that measures interference of pain on activities (e.g., school, sleep, mood, physical activities). Raw scores are converted to T-scores (mean = 50, SD = 10). Higher scores indicate greater pain interference. MCID = 3–7 points.
Pain Catastrophizing Scale – Child (PSC-C)24,38	13-item tool that measures excessive negative thoughts to actual or anticipated pain. Raw scores range from 0–52 with higher scores indicating higher degrees of catastrophizing. Cutoff scores have been determined in pediatrics: low (0–14), moderate (15–25), and high (>26) degrees of pain catastrophizing.
Tampa Scale for Kinesiophobia (TSK-6)26,39	6-item tool that measures fear of movement. Total scores ranging from 6–24, with higher scores indicating greater degrees of fear of movement.
Child Presence Measure (CPM)	12-item tool that measures subjective experience of presence and immersion in a virtual environment. Scores range from 0–24 with higher scores indicating higher degrees of immersion.
Patient Global Impression of Change	Single-item tool that measures perceived change in medical condition, rated on a 7-point Likert scale (1 = very much worse to 7 = very much improved; 4 = no change).
Child Treatment Expectancy/ Credibility (TEC-C)40	6-item tool that measures expectations and credibility of a treatment intervention. Each item is rated on a 0–10 NRS with higher scores indicating greater perceived treatment credibility and expectancy.
Numerical Pain Rating Scale22	Single-item tool that measures pain intensity rated on an 11-point scale (0 = no pain to 10 = worst pain imaginable).
Pittsburgh Rehabilitation Participation Scale21	Single-item tool that measures participation during therapy sessions, rated on a 6-point Likert scale (1 = none to 6 = excellent).
Modified Borg Rating of Perceived Exertion Scale – Dyspnea41,42	Single-item tool that measures breathlessness with exercise, rated on a 11-point NRS (0 = nothing at all and 10 = maximal).
Performance-based
6 Minute Walk Test (6MWT)32,33	Distance walked (meters) in six minutes under standardized conditions, longer distances indicate greater submaximal aerobic capacity and walking endurance. Normative values for pediatric population stratified by age and sex group.
Closed Kinetic Chain Upper Extremity Test34	Number of alternating touches completed between two 36-inch lines in 15s (average of three trials). Higher scores indicate greater upper-extremity stability and power.
30-Second Sit-to-Stand Test31	Number of full stands completed in 30 seconds from a 17-inch-high chair with arms crossed. Higher scores indicate greater lower-extremity strength and endurance.
Single Leg Balance Test43,44	Time (seconds) balancing on one leg with arms crossed until loss of position; longer times indicate better balance.
6 Minute Walk Test (6MWT)32,33	Distance walked (meters) in six minutes under standardized conditions, longer distances indicate greater submaximal aerobic capacity and walking endurance. Normative values for pediatric population stratified by age and sex group.

NRS = Numerical rating scale; MCID = Minimal clinically important difference, MDC = minimal detectable change, SEM = Standard error or measurement; SF = Short form.

Data collection, storage and analysis

Quantitative data from self-report, physiotherapist-report, and performance-based measures were collected and stored in REDCap.¹⁸ Data was analyzed using R statistical software (R Core Team, 2024, Version 4.4.1). First, data was inspected for missingness and errors, then, data was cleaned. Descriptive statistics (frequency, means, standard deviations [SD]) were used to summarize participant characteristics and measures of feasibility, acceptability, and effectiveness. Means (standard deviations) were reported for all measures except for highly skewed data where median (1st and 3rd quartiles) were more appropriate. As this was a pilot RCT and measures of effectiveness were exploratory only, we did not test for significant differences between intervention arms.

Actigraphy data were recorded by the Centerpoint Insight Smart Watch accelerometer (Ameritis) and was downloaded to Actilife software for processing. Data were processed for each of the four 1-hour treatment sessions (T1), and across the study period (between T1 and T2). Physical activity (PA) metrics included total number of minutes of PA and number of light, moderate and vigorous PA minutes.

Qualitative interviews were audio recorded and transcribed verbatim. Data was stored and managed in Dedoose software (SocioCultural Research Consultants, Version 0.0.17) and analyzed using qualitative content analysis.³⁵ Both inductive codes (drawing from participant transcripts) and deductive coding (drawing from the study design, protocol and aims) were used. Coding frameworks was developed by RX and GM using an iterative process; separate frameworks were created for the patient interviews and physiotherapists’ focus group. Transcripts were coded by RX and UK. Once the transcripts were coded, overarching categories were created to group related codes, and members of the research team (GM, RX, JS, TK) met to review and refine the category names and descriptions to align with the study objectives. Qualitative and quantitative data were analyzed separately; then, results were combined to develop insights across study objectives.

Sample size

As pilot RCTs are not powered for hypothesis testing, the sample size of 60 participants (20 per arm) was determined based on recommendations for feasibility trials from the CONSORT extension guidelines.¹⁶

Results

Thirty-seven participants were enrolled in the study, including 28 females (76%) and 9 males (24%). Participants were between the ages of 12–17 years and predominately identified as White (70%). Participant demographic characteristics, stratified by study arm, are presented in Table 2. Thirty-seven participants completed baseline measures (T0; n = 37), 30 participants engaged in 115 treatment visits (T1; n = 30) and 28 participants completed the study including endpoint measures (T2; n = 28). Ten participants engaged in semi-structured interviews and 4 physiotherapists engaged in the focus group.

Table 2.

Participant characteristics.

	Off-the-shelf VR N = 121	Custom VR N = 111	Control N = 141
Age (years)	15 (14, 17)	14 (12, 17)	16 (13, 16)
Sex
Female	8 (67%)	8 (73%)	12 (86%)
Male	4 (33%)	3 (27%)	2 (14%)
Race/Ethnicitya
Chinese	3 (25%)	0 (0%)	1 (7.1%)
South Asian	3 (25%)	3 (27%)	3 (21%)
White (Caucasian)	8 (67%)	9 (82%)	9 (64%)
Other	0 (0%)	1 (9.1%)	0 (0%)
Type of chronic pain
Headaches	2 (17%)	0 (0%)	0 (0%)
Abdominal pain	2 (17%)	1 (9.1%)	2 (14%)
Widespread musculoskeletal pain	4 (33%)	1 (9.1%)	5 (36%)
Low back pain	0 (0%)	2 (18%)	1 (7.1%)
Cancer pain	0 (0%)	0 (0%)	0 (0%)
Neuropathic pain	1 (8.3%)	0 (0%)	4 (29%)
Pelvic pain	0 (0%)	0 (0%)	1 (7.1%)
Facial Pain	0 (0%)	1 (9.1%)	0 (0%)
Other	3 (25%)	6 (55%)	1 (7.1%)
Pain duration (months)	19 (9, 36)	48 (8, 48)	25 (14, 72)

¹n (%); Median (Interquartile Range).

aother racial/ethnic categories where n = 0 (0%) included Filipino, South East Asian, Latin American, Korean, Japanese, Black, Aboriginal, Arab, West Asian, Prefer not to answer; other pain categories where n = 0 (0%) included Cancer pain.

Feasibility

Between June 2022 and October 2024, 738 patients were assessed for eligibility. However, 86% (n = 637) of patients did not meet the eligibility criteria (common reasons including geography, age, discharged from clinic, PT treatment not indicated, and visual, auditory or cognitive impairment). Of 70 patients who were deemed eligible and approached, 37 consented to participate in the study (enrollment rate 53%) and were allocated to a study arm (off-the-shelf VR [n = 12], custom VR [n = 11] and control [n = 14]). Recruitment was terminated before the target sample size was met as the funded study period came to an end, and the pool of eligible participants was exhausted. The drop-out rate was 33% (9 drop-out/37 enrolled). Reasons for drop-out included pain resolution (n = 1), no longer eligible due to medical/mental complexity (n = 4), and no longer interested/willing to participate (n = 4, including 3 who were allocated to the control group and withdrew prior to intervention as they expressed dissatisfaction with treatment allocation). Twenty-eight participants completed the study. Figure 2 CONSORT diagram outlines the flow of participants through enrollment to study end.

Figure 2.

CONSORT diagram.

VR = Virtual reality.

With respect to treatment visit adherence, 25 of 28 participants completed 4/4 sessions (3 participants missed a total of 4 sessions). With respect to completion rates of study measures, at baseline (T0, n = 37), all participants completed survey and performance-based measures (100% completion), at endpoint (T2, n = 28 participants) 100% completed performance-based measures and 96% completed survey measures. After every treatment visit (T1, n = 115 sessions), the patient survey completion rate was 90% (n = 104/115) and PT survey completion rate was 90% (n = 103/115 sessions).

Acceptability

Analysis of participant interviews, PT focus group and T1 and T2 survey data indicate that participants in both VR groups experienced high levels of satisfaction. One hundred percent of participants in the off-the-shelf VR group and 88% in the custom VR group were satisfied/very satisfied with the program. Participant and physiotherapist quotes from interviews are presented in Table 3, highlighting their perspectives on treatment acceptability. Treatment expectancy was similar across the three study arms. At baseline (T0), patient perceptions regarding the credibility of treatment (“does this therapy make sense to you?”) ranged from 7.1 to 8.1 (mean score) across three arms and potential effectiveness of the treatment (“how successful do you think this treatment will be in improving your functioning?”) ranged from 6.7 to 7.4 (mean score) across three arms.

Table 3.

Physiotherapist and patient perspectives on virtual reality feasibility and acceptability.

Category	Quote
Ease of use	“I thought it was easy to use generally. Not too many tech issues with it, and kids are usually pretty able to get it done at home.” – Physiotherapist #2
	“The only tricky piece was figuring out how to cast it … I remember emailing in a panic one time because I couldn’t get it to work.” – Physiotherapist #1
	“Anytime I don’t [use it], it is because I’m like, oh, it’s going to take me a long time to set it up and I really have a very short amount of time to treat them” – Physiotherapist #4
Fun and enjoyable	“It was good as well. I spent a lot of time on it when I could. It was fun because sometimes I would connect it to my TV and my mom could watch me play.” – 17 y.o. F
	“It was really fun because I’ve not really used VR before, so it was definitely a good experience. The games and everything you could do were very interesting.” – 13 y.o. F
	“It made physiotherapy a lot more fun, and I was able to do more activity.” – 17 y.o. F
	“Most kids enjoyed it. Families thought it was a cool new thing and were excited to use it.” – Physiotherapist #4
Distraction from pain	“My shoulder didn’t hurt as much because when I was playing a game … those are the things that help bring my pain down or bring my focus away from it.” – 12 y.o. M
	“It was really distracting. While I was playing, it distracted me from my pain, so that was a good thing.” – 15 y.o. F
Engagement in exercise	“I definitely got a lot of movement from the VR … I wouldn’t have been moving as much without it.” – 13 y.o. M
	“I think it maybe helped me move more because I was definitely working more and working harder … like more movement.” – 13 y.o. F
	“I had a lot of kids be super invested, and they were more inclined to do their home exercises if it involved VR.” – Physiotherapist #2
Targeting fear with movement	“I had one kid with CRPS who would never, ever put his foot on the ground – and with VR he could do half an hour of foot planting. It reinforced the power of distraction.” – Physiotherapist #1
	“I see myself using it for kids who need that major distraction to expose them to touch or movement – something new and scary.” – Physiotherapist #4
	“I used the VR to distract her because lying on her back was a big fear component … it helped her tolerate that position more.” – Physiotherapist #2
	“I felt like I didn’t worry as much about doing something that might make my pain worse.” – 13 y.o. F
	“When I was using [VR], I could put my foot down a bit and put more weight on it … after, I would still be a bit distracted so it was less painful. […] I don’t think I would have been able to do those movements if I didn’t have the VR.” – 12 y.o. M

y.o. = year old; M = Male; F = Female.

Ease of use: At endpoint (T2), 100% of participants in both VR groups rated the VR as ‘easy’ or ‘very easy’ to use. Some participants noted challenges with initial implementation for at home use (e.g., comfort with hardware, navigating software, following instruction guides) and interruptions in play due to issues with login, WIFI connection, glitches in the gaming software, or forgetting to charge the device. Participants reported these issues as minor and occurred infrequently.

Physiotherapists expressed several barriers that impeded the delivery of the intervention. With respect to logistical challenges, physiotherapists identified difficulty locating a suitable private clinical space from which to conduct videoconference sessions with participants via Zoom for Healthcare and to supervise the participant’s VR experience via screen cast. Physiotherapists were also responsible for device care between sessions, including wiping down the VR headset, controllers, face cover, and accessories with a virucidal and bactericidal hydrogen peroxide solution (Virox™), followed by a 4-minute UV-based disinfection cycle using the Cleanbox CX1, in accordance with hospital’s Infection Prevention and Control guidelines, as well as charging and storing devices between use. When devices were stored in less accessible areas or appointment time was limited, setup felt burdensome as outlined in Table 3.

With respect to technical challenges, physiotherapists reported difficulties with casting, managing shared device accounts, and syncing devices with Zoom for Healthcare for remote supervision, as well as WIFI connectivity. Physiotherapists expressed that study staff were instrumental in overcoming both logistical and technical barriers by providing real-time support when needed to minimize interruptions in care.

Enjoyment: At endpoint (T2), 88% of participants in both VR groups ‘enjoyed’ or ‘very much enjoyed’ using VR during the physiotherapy session. Further, at the final PT session (T1), participants in the VR groups reported higher average levels of enjoyment (mean scores [SD]: off-the-shelf VR 66.9 [21.9], custom VR 76.8 [27.0] out of a total possible score 100) compared to participants in the control group (n = 11, mean score [SD]: 54.5 [23.7]). Participants and physiotherapists expressed that VR was fun and provided a powerful distraction from pain while doing exercise. Participants in the VR groups reported higher average levels of distraction (mean scores [SD]: off-the-shelf VR 53.0 [32.7], custom VR 56.3 [17.7] out of a total possible score 100) compared to participants in the control group (n = 11, mean score [SD] 28.9 [16.1]).

Engagement during each PT session: Participants in both VR groups reported similar levels of immersion on the Child Presence Measure, where mean scores [SD] in the off-the-shelf group were 17.4 [3.9] and custom VR group 20.0 [3.2] out of a total possible score 24. Across all three study arms, participants reported moderate levels of exercise intensity on the Borg Rating of Perceived Exertion scale where mean scores in the off-the-shelf group were 3.5 [2.9], custom VR group 4.9 [1.8] and in the control group 3.7 [2.1], out a total possible score of 10 (scores >7 indicate high intensity.³⁶) With real-time exercise monitoring using Actigraphy, the total number of minutes of PA (mean [SD]) were similar between the VR groups (44 min [24]) and the control group (38 min [23]), although the number minutes of moderate-vigorous PA (mean [SD]) were highest in the VR groups (off-the-shelf 20 min [23], custom VR 28 min [33], control group (10 min [13]). Across all three study arms, physiotherapists reported similarly high levels of engagement on the Pittsburgh Rehabilitation Participation Scale where mean scores [SD] were 5.1 [1.1] off-the-shelf group, 5.0 [0.7] in custom VR group, and 4.9 [1.3] in the control group, out of a total possible score of 6.

In participant interviews, participants expressed that the gaming nature of VR made exercise feel less burdensome and allowed them to do specific movements they would typically avoid and/or do overall more exercise than they normally would. Participants expressed that compared to other types of PT, they preferred VR as it was more fun, motivating and engaging and this made them more willing to complete PT exercises. Physiotherapists also perceived that VR motivated participants to engage in exercise, especially exercises that were fearful or avoided. Physiotherapists noted that patients achieved higher levels of mobility, weightbearing, or exercise intensity than expected when using VR as opposed to other types of PT.

Safety

Ten minor adverse events were reported by physiotherapists during the virtual PT sessions (7 in the VR study arms and 3 in the control group) involving 8 participants. Two additional adverse events were reported by participants related to the actigraphy watches (contact dermatitis that resolved with discontinued use). Events reported with VR use included dizziness (n = 1), feeling disoriented (n = 2), fatigue (n = 1) and pain (n = 3); these events resolved with dismounting the headsets. During participant interviews, some participants reported temporary side effects with VR use at home, including dizziness, nausea, headaches, and feeling disoriented, after longer or more intense gameplay.

Exploratory measures of effectiveness

Participant results on measures of effectiveness, including self-report and performance-based measures, are displayed in Table 4.

Table 4.

Secondary measures of effectiveness.

	Off-the-shelf VR		Custom VR		Control
Measures	Baseline N = 121	Endpoint N = 81	Baseline N = 111	Endpoint N = 91	Baseline N = 141	Endpoint N = 111
Pain intensity
NRS – Average	5.3 (1.4)	5.4 (1.8)	6.5 (1.8)	5.6 (2.3)	4.9 (2.1)	3.3 (2.1)a
NRS – Current	5.3 (2.0)	6.3 (0.7)	6.3 (2.1)	5.4 (2.5)	3.4 (2.4)	3.0 (2.3)a
Psychological constructs
FOPQ – Total Score	32.8 (5.4)	31.4 (4.2)	33.9 (7.9)	29.9 (8.1)	32.9 (8.8)	30.0 (8.2)a
PCS-C – Total Score	27.1 (11.6)	22.0 (7.9)	30.8 (13.2)	29.2 (8.7)	27.4 (13.6)	22.2 (8.2)a
TSK-6 – Total Score	13.5 (3.8)	13.9 (3.1)	15.0 (4.9)	15.0 (4.3)	14.8 (4.6)	12.4 (3.0)a
Physical function
UEFI – Total Score	57.6 (22.1)	50.6 (25.2)	57.1 (18.1)	59.0 (20.1)	63.4 (12.1)	58.4 (15.6)a
LEFS – Total Score	45.7 (13.8)	42.5 (18.9)	44.0 (19.2)	48.8 (17.3)	50.7 (16.9)	48.7 (17.1)a
PROMIS Mobility V2.0 – T-score	36.3 (4.0)	33.6 (6.0)	34.6 (9.5)	36.9 (9.7)	39.9 (8.8)	38.4 (8.9)a
PROMIS Pain Interference V1.0 – T-score	63.9 (7.0)	61.8 (5.2)	65.7 (5.8)	60.3 (8.5)	62.7 (6.8)	59.7 (6.4)a
6MWT – meters	421.3 (176.7)	429.0 (255.1)	288.0 (140.9)	375.1 (90.6)	405.1 (123.8)	392.7 (83.9)
Sit to Stand	14.8 (5.9)	12.6 (5.0) a	11.0 (2.0)	12.4 (6.5)	11.5 (3.2)	13.5 (3.8)
Closed Kinetic Chain Upper Extremity Test	14.3 (6.9)	14.7 (3.9) b	13.9 (5.2)	12.6 (6.9)	11.9 (4.7)	13.6 (4.4)
Single leg stance – eyes open (median [IQR])	60 (53, 60)	60 (36.5, 60)	60 (60, 60)	60 (60, 60)	60 (60, 60)	60 (46, 60)

¹Mean (Standard Deviation); ^aMissing n = 1, ^bMissing n = 2.

NRS = Numerical Rating Scale; FOPQ = Fear of Pain Questionnaire’ PCS-C = Pain Catastrophizing Scale Child; TSK-6 = Tampa Scale Kinesiophobia 6-item short form; UEFI = Upper Extremity Functional Index; LEFS = Lower Extremity Functional Scale; PROMIS = Patient-Reported Outcome Measurement Information System; 6MWT = 6 Minute Walk Test; IQR = Interquartile Range.

Pain: At baseline and endpoint, participants in all study arms reported moderate levels of pain intensity (mean scores on the NPRS ranged from 3.3 to 6.5/10 across groups) and pain interference (mean scores on the PROMIS Pediatric Pain Interference scale ranged from 60 to 66/ 100). For participants in the VR groups (n = 16), 38% reported that VR reduced their pain “a medium amount” and 13% “very much.” In interviews, participants described that although VR did not change their pain, they valued that VR provided a distraction from pain to allow them to exercise.

Fear and catastrophizing: At baseline and endpoint, participants across study arms reported moderate (15–25) to high (>26) levels of pain catastrophizing on the PCS-C where mean scores ranged from 22 to 31 out of a total possible score 52. At baseline and endpoint, participants across study arms reported similar levels of fear of movement, with mean scores on the TSK ranging from 12.4 to 15.0 out of a total possible score 24.

Physical function: With respect to mobility, mean scores across groups on the PROMIS® Pediatric Mobility scale were similar at baseline (35–40) and endpoint (34–38), indicating moderate impairments in mobility (scores range 0–100). With respect to upper and lower extremity function, mean scores across groups at baseline (UEFI: 57–63; LEFS: 44–51) and endpoint (UEFI: 50–59, LEFS: 43–49) were similar (scores range 0–80). With respect to walking distance, participants’ scores at baseline and endpoint were below the normative values for youth 9–16 years (540–865 m). Participants in the custom VR group had the greatest improvement in number of meters (m) walked from baseline to endpoint on the 6MWT (off-the-shelf VR +7 m, custom VR +87 m, and control group −12 m). With respect to real-time monitoring of exercise using Actigraphy, across the study period (T1–T2), the average daily number of minutes of light PA (mean [SD]) were similar between the VR groups (268 min [113]) and the control groups (267 min [89]), although the average number of minutes of moderate PA were higher in the VR groups (18 min [22]) compared to control group (10 min [6]).

Global impression of change: At endpoint (T2), participants across each study arm reported similar levels of overall change on the PGIC measure (Fig. 3), with mean scores [SD] in the off-the-shelf 3.4 [1.2], custom VR 2.8 [1.0] and control 3.0 [1.3], out of a total possible score 7.

Figure 3.

Patient global impression of change score at study end (n = 28).

VR = Virtual reality; missing n = 1.

Response options range from 1 = very much worse, 2 = much worse, 3 = minimally worse, 4 = no change, 5 = minimally improved, 6 = much improved, and 7 = very much improved.

Missing data

There were less than 10% missing data for satisfaction surveys and secondary measures of effectiveness; this was addressed using complete case analysis.

Discussion

This pilot randomized controlled trial is the first study to evaluate pain rehabilitation virtual reality (PR-VR) for at home use as part of physiotherapy (PT) exercise-based treatment for youth with chronic pain. This was a 3-arm clinical trial comparing two types of VR software incorporated into PT (off-the-shelf VR software and custom-developed PR-VR software) to standard PT. The primary aims were to evaluate the feasibility, acceptability and safety of PR-VR for youth with chronic pain. An exploratory aim was to examine the preliminary effectiveness of PR-VR on pain and disability. Study findings indicate that PR-VR is a safe and acceptable form of physiotherapy treatment to both patients and physiotherapists. Challenges encountered in this feasibility study will inform the design of a larger-scale clinical trial.

Comparison to existing literature

Virtual reality is an emerging technology in the management of chronic pain with a larger body of evidence in adult populations compared to pediatric population. The vast majority of the existing pediatric VR literature focuses on acute pain, rather than chronic pain populations. This study is one of a few studies targeting VR for pediatric chronic pain and is the first study to examine the remote delivery of VR for this population. A systematic review of extended reality (which include VR, as well as augmented reality) interventions found VR in the pediatric acute pain setting is feasible and safe, in alignment with the current study.⁹ In adult chronic pain setting we find generally high adherence rates, minimal adverse events, and strong engagement with at-home VR interventions, with some studies also demonstrating sustained improvements in pain and function following treatment.^37–40 This contrasts with the present study where some important feasibility challenges were identified. These differences may be a result of differences in study designs, the type (active, waitlist) and number of control conditions and samples.

Feasibility

The metrics pertaining to study feasibility had mixed findings. Although there is no well-established threshold for acceptable enrollment rates, the enrollment rate in this study (>50%) was considered excellent given the vulnerable nature of the population (youth with pain) and the nature of the intervention was relatively high demand (including daily exercise programs). Similarly, the completion rates for study measures were also excellent (>90% completion rate for study surveys and performance-based measures). For participants who completed the study, adherence to study visits was excellent (>90%).

Several challenges encountered in this trial will be used to inform the design of future trials. First, 86% of patients screened for the study were deemed ineligible. In this study design, the eligibility criteria were stringent; reexamination of inclusion/exclusion criteria is warranted. For example, many patients were deemed ineligible due to distance from the hospital (>150 km) as the study design required pick-up/drop-off of study equipment. However, given the remote nature of the study intervention, this criterion could be eliminated and study equipment could be shipped to/from study participants. This approach has been successfully implemented in an adult chronic pain population.³⁹ This would not only facilitate study enrollment but also enhance access to this remote treatment.

Second, the drop-out rate was high (33%). One-third of dropouts occurred after enrollment, before the study intervention, because participants were dissatisfied with allocation to the control group as they enrolled in the study hoping to use VR. Given the nature of the intervention, blinding participants to the intervention is not possible. Future trials could consider alternate designs to avoid a traditional control group. For example, using a wait-list control or within-subject design so that all participants receive a VR intervention, or a pre-post interventional trial without any control group.

Third, we were only able to enroll two-thirds of the desired sample size at our single study site. Future larger trials may require multi-center collaborations and consider a 2-arm vs 3-arm design. Merging the two VR groups (off-the-shelf VR and custom VR) into a single VR group would allow study participants to have a larger menu of possible software games to select from. This has several advantages, including allowing patients/physiotherapists to select games based on their interest and clinical needs and to allow for a variety of games within session and across the study period to sustain patient interest. Additionally, future studies should consider incorporating camera-based hand tracking VR environments (e.g., MetaQuest2), particularly for pediatric upper limb CRPS patients requiring hand therapy, as this approach may enhance avatar ownership which may improve treatment outcomes.³⁷ With respect to clinical needs, physiotherapists indicated PR-VR games were particularly helpful for patients with single-limb pain with a low-level of physical function, and commercially available software were appropriate for patients with widespread pain and/or higher levels of physical function. With respect to sustaining interest over time, patients with chronic pain may require long-term therapy and/or recurrent episodes of therapy, and exercise is ideally sustained after completion of therapy as part of a healthy lifestyle. Beyond PT, VR technology also holds promise for other pain interventions such as mindfulness and distraction-based approaches.⁹

Acceptability

Both patient participants and physiotherapists expressed high levels of acceptability. Patients expressed that VR was fun, enjoyable, and provided a powerful distraction from their pain experience. Both patients and physiotherapists reported that VR was helpful to better engage patients in exercise. For example, by motivating them to initiate exercise at home, exercise for longer periods, and/or participate in exercises that were feared and avoided. Patient participants reported minimal technical issues, and physiotherapists expressed although technical issues did arise, study staff were able to address them in a timely manner to minimize the impact on clinical care. These findings align with findings from previous research of PR-VR in the day-treatment environment.¹⁰

Safety

This trial indicates VR is a safe technology to engage youth with chronic pain in PT exercises. Although some participants reported adverse events with VR, these were mild and transient. Additionally, the number of adverse events reported in the VR groups (7 events in a group of 16 participants) were similar to the control group (3 events in a group of 11 participants). This suggests that although there may be mild discomfort experienced while engaging in exercise using VR, this is similar to traditional PT exercises.

Effectiveness

As an exploratory aim, we sought to generate a preliminary estimate of effectiveness by describing participants baseline and endpoint metrics of pain and disability. As this was a pilot study, the sample size was not sufficient to detect between-group differences. At baseline, with respect to pain, patients had moderate levels of pain intensity and pain interference, and with physical function, patients had moderate-level impairments in mobility and below-average walking endurance. Participant scores at endpoint appeared similar to baseline across all study groups. There appeared to be some differences across baseline to endpoint in walking endurance and pain interference. Future clinical trials may consider a higher dose of the intervention, as 4 treatment sessions is relatively low dose for exercise interventions, especially for patients with a chronic condition.

Study strengths and limitations

This study has several notable strengths, including examining an understudied intervention (physiotherapy) using novel technology (VR) in a vulnerable population (youth with chronic pain). The study design was robust in accordance with the SPIRIT guidelines.¹⁷ Although blinding physiotherapists and participants to the study intervention was not possible, study outcome assessors were blinded to minimize risk of bias. Survey completion rates were excellent and therefore missing data (<10%) was not felt to compromise the validity of results. There are also several study limitations worth noting that impact the generalizability of results. First, recruitment closed at the end of the funded study period with a sample size of 37, below the recommended threshold for pilot feasibility trials (n = 20 per group, total n = 60), which may impact the validity of the safety, feasibility and acceptability results. This was largely due to the study design (single-center) with a reduced pool of eligible participants related to restrictive inclusion criteria. Future trials should consider multi-center recruitment and less restrictive inclusion criteria to increase the pool of eligible participants. Second, the sample was largely Caucasian (70%); although this is representative of the racial distribution of patients in the recruiting clinic, this is unlikely representative of the racial distribution of youth living with chronic pain in the general population. Study results may be less generalizable to nonwhite youth with chronic pain. Third, although the study sought to include children and youth (8–18 years), participants were youth 10–18 years and study results may not be generalizable to children <10 years old. Fourth, there were some baseline differences between groups (e.g., lower pain levels in control group) future trials should address this by using stratified randomization. Fifth, avatar embodiment (i.e., the extent to which users feel that a virtual body represents their own body) was not measured and may be an important factor to consider in future studies, particularly when incorporating more immersive features.

Conclusion

This pilot study demonstrates the feasibility, acceptability and safety of using VR to engage youth with chronic pain in exercise-based physiotherapy in the home environment. Successes (safety, enrollment rates, high levels of patient satisfaction) and challenges (high dropout rates) experienced during this trial will inform the design of future larger-scale clinical trials to determine the effectiveness of this intervention.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Funding

This work was supported by Physician Services Incorporated Foundation under the Resident Research Grant [Grant #2020-1864] and Johnson and Johnson under the Nurses Innovate QuickFire Challenge in COVID-19 Patient Care grant [Grant #0000000393].

Data availability statement

The data that support the findings of this study are located in controlled access data storage at the SickKids Research Institute. This data is not publicly available as the participants of this study did not give written consent for their data to be shared publicly.

Supplementary Material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/24740527.2026.2683648