The Immersive Cleveland Clinic Virtual Reality Shopping Platform for the Assessment of Instrumental Activities of Daily Living

The CC-VRS Platform, to date, appears to most effectively address the locomotion problem in VR by combining state-of-the-art VR content with an omnidirectional treadmill. A critical aspect of the seamlessly immersive environment of the CC-VRS is ensuring proper communication between the treadmill and VR software. Correct setup of all aspects of the VR system—including base stations, feet and waist trackers, and hand controllers—is imperative. If tracking is inconsistent or unreliable, adjustment of the orientation and placement of the base stations, or addition of another base station unit, is required. Proper coverage of the physical space provides stable synchronization between the VR hardware and omnidirectional treadmill and ensures the position and orientation data from the VR devices are complete, accurate, and precise⁵². Calibration of the omnidirectional treadmill is recommended at the start of every data collection session to ensure optimal responsivity while navigating the virtual environment.

Patient familiarization with the treadmill is critical prior to administering the CC-VRS. Although intuitive and simple to operate, the omnidirectional treadmill requires user familiarization that is best done prior to introduction of the VR headset and the resultant spatial orientation challenges. To meet the needs of the individual user and the goals of the present assessment, the following features are configurable for each CC-VRS scenario: 1) treadmill low or high maximum speed, 2) repetition of tutorial module, 3) route length through the store, and 4) number of items on shopping list. These modifications optimize the assessment for the cognitive, motor, and dual-tasking capabilities of a wide functional range of patients.

The lack of single-platform technology capable of standardizing IADL performance by using objective and quantitative outcomes that characterize cognitive and motor functioning represents a critical barrier in the early identification and effective treatment of age-related neurological diseases such as PD or Alzheimer's disease. Current methods estimating IADL function using self-report questionnaires, while easy to administer, are susceptible to bias. During self-report, older adults tend to over- or under-estimate IADL capabilities⁵³. Similarly, informants completing IADL queries often misjudge capabilities due to the observers' misperceptions or knowledge gaps³⁵.

An alternative to self-report and informant-rated questionnaires is performance-based IADL evaluation. Performance-based evaluations are typically completed by a trained Occupational or Physical Therapist. While a number of performance tests and guides are available, they are not conducive to integration into clinical care, often requiring ample time and specialized space and equipment not typically found in a primary care or neurology provider's office. One of the most widely used performance-based assessments, Direct Assessment of Functional Status (DAFS), requires approximately 40 min to administer, and its scoring is largely based on expert opinion of the test administrator. Although the DAFS is useful in staging Alzheimer's disease, it lacks sensitivity and does not detect IADL declines at the mild cognitive impairment stage²⁴. Fusing the virtual and real worlds by combining VR with an omnidirectional treadmill provides an opportunity to capture IADL performance under complex cognitive conditions that better replicate real-world environments, potentially resulting in earlier diagnoses of neurological disease⁵⁴.

The CC-VRS platform addresses the clinical gap by providing a standardized, systematic, objective, and quantitative approach to characterizing IADL capabilities in older adults and those with neurological disease. Based on preliminary usability testing and data, the Basic and Complex CC-VRS Scenarios can be completed altogether in less than 30 minutes. Similar to other immersive VR studies with PD¹⁸, the majority of people with PD experience mild motion sickness symptoms. From a usability perspective, the CC-VRS received an overall SUS rating of 75.7, corresponding to a letter grade "B" and falling between the "good" and "excellent" descriptor categories⁵⁵. For comparison, a recent assessment of popular phone and tablet applications reports an average SUS score of 77.7 for the top 10 applications across all platforms⁵⁶, including mobile applications such as The Weather Channel and YouTube. Comments from participants indicated that most users enjoyed the realism and ability to interact with the CC-VRS platform. Importantly, the participants felt challenged from a physical and cognitive aspect, indicating the design achieved its goal to create a dynamic platform that recreated a complex IADL experience.

We have previously demonstrated that technology can be successfully integrated into clinical workflows in the evaluation of patients with concussion⁵⁷ and in a busy neurological service for patients with Multiple Sclerosis (MS)⁵⁸. Further, the use of technology in the management of concussion improved outcomes and reduced costs⁵⁹, while its use in the treatment of MS led to a 27% saving in time spent documenting in the electronic health record for each patient⁶⁰. Considering the continual goal to reduce the cost of delivering care⁶¹ and that time spent documenting in the electronic health record is frequently cited for physician burnout⁶², the integration of the CC-VRS platform into clinical care is likely to provide a substantial value-add to hospital systems. Two projects are ongoing in which the CC-VRS platform is integrated into 1) a regional primary care family health center that primarily treats healthy older adults and 2) a specialized movement disorders clinic at the Cleveland Clinic.

The absence of an accurate and reliable physiological or digital biomarker for PD and Alzheimer's disease causes great difficulty in early diagnosis and in measuring disease progression^63,64. The CC-VRS platform has the potential to provide a digital biomarker under a single technological platform that will enhance clinical care and could result in shorter and more efficient clinical trials by reducing reliance on subjective and highly variable clinical outcomes (e.g., Movement Disorder Society – Unified Parkinson's disease Rating Scale motor portion (MDS-UPDRS III)). The evaluation of motor and cognitive function in the field of clinical neurology has not advanced dramatically over the past three decades in terms of the assessment of individuals with PD and the associated cardinal motor symptoms, let alone cognitive or dual-task problems. The most celebrated advancement in the assessment of individuals with PD is the revision of the subjective clinical rating scale (MDS-UPDRS III). Importantly, we do not believe that the CC-VRS will supplant the MDS-UPDRS III. Rather, we believe its greatest value may be realized in primary care practices by providing a standardized and objective approach to the quantification of IADLs. While it is premature to believe the CC-VRS in its current form is a prodromal marker of neurological disease, results could be used to raise a "red" or "yellow" flag in terms of neurological functioning that may trigger a consult by a movement disorders, neuropsychology, or geriatric specialist. In terms of its use in PD clinical care, it is anticipated that the CC-VRS could be utilized in the titration of medication or in the eventual programming of deep brain stimulation devices. Both the Primary Care and PD-specific use cases are currently in the pilot phase. By truly immersing the user in a realistic environment and measuring meaningful and important aspects of cognitive and motor function, the CC-VRS represents an initial step in the creation of a potentially effective and scalable digital biomarker for neurological disease.

The field of clinical neurology, movement disorders in particular, is filled with examples of technology developed to quantify a single, isolated PD symptom via accelerometer or other sensor technologies^{65,66,67,68,69}. To our knowledge, none of these approaches, other than our balance^70,71,72,73 and tremor applications⁷⁴, have been integrated into routine PD clinical care. Previous technology oftentimes is valid and reliable; however, the focus has been on technology development with little regard to feasibility of clinical integration^75,76. Patients, providers, hospitals, and regulatory bodies are increasingly interested in outcome measures that quantify changes in meaningful daily actions^77,78,79,80. The clinical integration of precise and meaningful measures of neurological symptoms and IADL performance is necessary to systematically evaluate the overall effectiveness of an intervention or determine the potential of an intervention to slow disease progression. The development of a standardized approach to IADL assessment suitable for routine clinical use is appealing to facilitate comprehensive understanding and treatment of neurological disease on meaningful activities.

The CC-VRS approach to the evaluation of IADL performance to aid in the diagnosis and management of neurological disease has the potential to transform healthcare through early diagnosis and more precise tracking of disease progression. However, it is fully acknowledged that the system is not without limitation. The cost of the omnidirectional treadmill is substantial and may serve as a barrier for widespread adoption sans systematic health economics studies to identify the potential "break even" point between cost of the assessment relative to the value of early diagnosis or more precise tracking of disease progression. Notably, gaps in acquiring PD patient-centric outcomes with technology were highlighted by the National Institute of Neurological Disorders and Stroke PD Conference⁷⁸, MDS Task Force on Technology⁷⁷, and FDA. They called for technology that measures meaningful PD activities and integration of these outcomes into clinical care. We are currently evaluating the integration of the CC-VRS into a primary care setting and a movement disorders center at the Cleveland Clinic; these deployments will utilize more affordable omnidirectional treadmills. Successful collection of data does require an initial investment of time by the clinician to learn how to set up and operate the system. Ongoing clinical pilots will better inform the amount of training required to become a proficient user. One could imagine a model in which a technician is employed to operate the system, and patients complete the shopping tasks rather than sitting in a waiting room before an appointment. Those data could then be instantaneously integrated into the electronic health record prior to seeing their provider. These types of applications have the potential to become the waiting room of the future for patients.