Effect of Peripheral Vestibular Disease Location on Outcomes Following Home-based Virtual Reality Vestibular Therapy

Purpose:

To determine if the location of the peripheral vestibular lesion(s) impacts the efficacy of adjunct at-home head-mounted display virtual reality therapy to improve patient symptomatology 0-3 months after the intervention. Also to determine the adverse effects caused by at-home head-mounted display virtual reality therapy.

Justification:

Home-based head-mounted display virtual reality video game exercises utilize a virtual environment to encourage wearers to complete head movements that mimic vestibular rehabilitation exercises. This virtual reality intervention could act as a valuable adjunct to standard virtual reality therapy. However, current literature lacks information on how the location of peripheral vestibular pathology impacts the efficacy of HBVR vestibular rehabilitation. Adding this information to the field could further personalize vestibular rehabilitation to the patient. Our research study personnel, specialist clinic, and access to vestibular investigations positions us well to undertake this work.

Research Design:

This is a parallel-group randomized controlled trial involving multiple clinical visits (the exact amount depends on diagnosis).

Statistical Analysis:

A sample size of 40 (20 per group) was determined to be sufficient for this study based on previous literature and the desired power that we hope to achieve. A meta-analysis completed by the research team, which is currently being considered for publication, indicated that the mean of the standardized mean DHI differences in current randomized controlled trials on VR in vestibular rehabilitation is -1.13. One article from this meta-analysis with a similar methodology to this RCT had a standardized mean ABC difference of 1.0. Additionally, well-designed randomized controlled trials set their type II error rate for 0.20 or 0.10 which results in a power of 80 or 90 percent. Inputting these standardized mean differences and 90% power values into the nomograph published by Altman 1982, yields a maximum sample size of 30 for DHI and 40 for ABC. Adopting the larger sample size and making an allowance for an acceptable attrition rate of 20%, it is proposed to recruit 50 patients in order to achieve a final sample size of 40 and thus a power of 90% for DHI and ABC data.

Mean and standard deviations will be calculated for each group's DHI and ABC scores 2 weeks before and after the intervention. Mean and standard deviations for patient SSQ scores will be calculated for each week of the protocol. An unpaired two-sample Wilcoxon test and a paired two-samples Wilcoxon test will be used to determine how significant the between (control vs. experimental) and in group (pre vs. post-treatment) differences are respectively. Subgroup analyses will be conducted on three subgroups of patients within each control and experimental group. These subgroups will be based on lesion location: lateral semicircular canal (Group A), utricle (Group B), and saccule (Group C). Matching subgroups from control and experimental groups will also be analyzed using unpaired two-samples Wilcoxon test and a paired two-samples Wilcoxon test to determine how significant the between (control vs. experimental) and in-group (pre vs. post-treatment) differences are respectively. Statistical significance cut-off (α) will be set at a p-value of 0.05.

Project Flow:

The patient will be contacted by phone or UBC Zoom conference call two weeks before the intervention begins to attain baseline data. This will involve assessing patient symptomatology by completing the dizziness handicap inventory (DHI) and activities-specific balance confidence (ABC) questionnaires remotely. Next, an interview will be conducted on this call to collect additional patient information including age, sex, ethnicity, physical activity level and virtual reality experience. Lastly, the VR headset will be mailed to the patient's home address using insured FedEx shipping. A date (before intervention start) for a VR device tutorial will be discussed at this visit.

Next, patients will be divided into two groups using a computer-generated randomization schedule. These groups will both undergo standard vestibular rehabilitation administered by a registered physiotherapist. This is considered standard of care and requires 4 to 8 weeks of vestibular rehabilitation for unilateral vestibular hypofunction (UVH) and bilateral vestibular hypofunction (BVH) respectively. This entails once a week 40 to 45-minute in-person sessions with a registered physiotherapist and 3 sessions a day of at-home exercises. The in-person sessions involve a combination of 4 different exercises: (1) exercises to promote gaze stability, (2) exercises to habituate symptoms, (3) balance and gait training, and (4) aerobic exercise. Gaze stability exercises that promote adaptation and substitution are both performed. Adaptation gaze stability exercises involve head movement while maintaining focus on a stationary or moving target. Substitution gaze stability exercises could include eye-head exercises where you conduct a large eye movement to a target before moving your head to face the target. Habituation exercises are conducted to expose the patient to dizziness-provoking stimuli in order to reduce symptoms to these stimuli over time. This could involve the use of optokinetic stimuli or virtual reality environments. Balance and gait training exercises are used to facilitate the use of visual and/or somatosensory inputs to substitute for lost vestibular function. Balance exercises involve balancing whilst experiencing altered visual input (ex. eyes closed) and/or somatosensory input (ex. foam or moving surface). Lastly, aerobic exercise, such as a stationary bicycle, is used for general conditioning because patients with vestibular hypofunction tend to limit physical activity to avoid symptom provocation. The registered physiotherapist providing this therapy will be blinded to patient group allocation. Patients will also complete 3 sessions a day of at-home adaptation, substitution, habituation, and balance exercises for a total of at least 20 minutes per day. The exact exercises will be similar to the in-person sessions however they will be truncated and adapted to the home setting.

Additionally, for 20 minutes per day for 4 (UVH) or 8 weeks (BVH) the virtual reality group will utilize a head-mounted display device to play a video game projected on an android or apple device accommodated into a VR headset. At-home exercises and VR usage will be recorded by patients using a study diary. Parameters recorded in the diary will include VR intervention usage (time of day, duration), physical activity (time of day, duration), and diet (breakfast, lunch and dinner contents and time of day). The patient will be instructed on how to size and use the VR headset during a tutorial scheduled after the initial assessment. A simulator sickness questionnaire (SSQ) will be completed every day after either the home exercise program (control group) or VR game intervention (experimental group). Two weeks after the intervention period patients will complete a subsequent remote assessment of symptomatology (DHI and ABC) and will return the VR headset through an insured FedEx shipment to Dr. Nunez's Neurotology clinic at the Gordon and Leslie Diamond Health Care Centre in Vancouver BC. VR headsets will be disinfected according to Vancouver coastal health protocols prior to providing the headset to subsequent patients. Additionally, both study groups will have submitted their study diary and SSQs to the research team via email or mail by this time.

Prior to completing the above interventions, the video game used must be congruent to current vestibular rehabilitation program exercises taught by a Canadian university physiotherapy program. The exact video game has not yet been chosen, however, point-of-view racing games compatible with both Android and iOS devices will be evaluated. This is due to Micarelli et al. 2017 and Micarelli et al. 2019 demonstrating the symptomatic benefits of this class of VR videogame. This evaluation will require a member of the research team to be filmed playing various point-of-view VR racing games. These videos will be sent to faculty and students at a Canadian University's Physiotherapy Program. After watching the videos, these individuals will complete a survey with 5-point Likert questions that ask for their opinion on if the video game effectively mimicked standard vestibular rehabilitation. The video game with the highest average score across all faculty members and students will be used by participants within the virtual reality intervention group.