Challenges for clinical VR in remote patient monitoring research

By John Christy Johnson

As virtual reality (VR) engineers strive to produce more immersive experiences for users, it provides an appealing distraction, permitting people to, “lose”, themselves in the world of VR. People are so engaged that they do not notice what’s happening peripherally in their surroundings and may lose track of the time that has passed. One application that this immersion can largely benefit is in the context of monitoring patients in rehabilitation environments. In this setting, patients will be able to experience and pursue pleasures while researchers and clinicians obtain the relevant data they need. For example, quite recently, a tech company called VRHealth launched their new artificial intelligence (AI) trainer called Luna ―a cognitive behaviour therapy (CBT) programme for pain management in seniors. Through its VR-directed programming, Luna will be able to, “help seniors receive therapist-guided physical therapy without needing to leave their home”. While theoretically, this sounds like an exceptional idea―this scenario is fraught with problems. In this article, we will be considering some of the ulterior motives and issues that ought to be considered alongside this exciting development. Before entering this discussion, there ought to be a disclaimer that there will be a large degree of speculation involved.

Researcher and corporate agendas

“VR in remote patient monitoring is an exciting development and one that should be carefully thought about prior to widespread adoption”

Firstly, there is a likelihood that there is a conflict of interest for researchers involved with developing new VR technologies. As developers and knowledge leaders in VR are eager to push such technologies forward, several ethical concerns come into play. The researcher may push their own agendas to further research while disregarding patient autonomy. For example, a huge part of CBT and other exposure therapies are that these strategies are reliant on the patient confronting specific alterations that are introduced as stimuli. Perhaps this is best illustrated by work done in VR exposure therapy for patients with post-traumatic stress disorder (PTSD). The problem of VR in PTSD treatment, along with phobias and addictions, are that participants or patients of the VR programme are subjected to perceived harmful stimuli. As such, it can worsen symptoms of their phobias and addictions, should they be re-exposed to it with enough realism. Furthermore, the concept of informed consent becomes blurred. As researchers may choose to take decisions for them, the patients or participants are prevented from opportunities to provide informed consent and withdrawal from participation. Similarly, with independent companies commercializing VR technology for remote rehabilitation, a new level of partnership must be reached as hospitals, clinics and educational institutions rely on corporate support. This may mean that certain services and products will only be provided if medical professionals promote them to patients.

Lack of supervisory decision-making

Secondly, a lack of human feedback can be harmful, especially considering the previous example of CBT. The fact that clinicians are hampered from interfering with a programme that is largely machine-driven can mean that they would not be able to step in and withdraw the intervention provided if there are cues that indicate a participant cannot participate. While it can be argued that AI can make this decision for them using cues within the system―it may be hard for the patient or family caregiver at home to perceive when it is necessary to disengage from VR.

“The patient is denied the ability to socialize and receive more sensitive counsel from a physical human therapist”

Similarly, during this research period, the AI is trained with a plethora of stimuli that will be provided by sample patients. If there are rare, specific patient-centred needs that arise, it would take a longer period to train. We cannot ignore that there is a risk that patients may be given directives that actually lead to more harm than benefit.

Replacing healthcare jobs

The idea that you can have a virtual physical therapist that can supersede human ones inherently leads to the question: Why do we even need human physical therapists in the first place? This is a bold statement and perhaps it will take years or decades to reach this level, but it still ought to be thought about. In this case, perhaps inevitably it’ll lead to a decrease in specialized professions like this one as knowledge becomes offloaded onto servers and the cloud. With the loss of such specialties, perhaps novel types of professions will also effervesce forth. As these types of programmes are being developed, more engineers and technical skill-oriented professions will be needed to support the complex demands of the patient.

Loss of social presence

With the lack of physical therapists, there is a likelihood that patients are not given the same level of encouragement or human elements of interaction. For example, face-to-face communication would be replaced with avatars and audio prompts in headsets. Regardless of how realistic the VR may be, the patient is denied the ability to socialize and receive more sensitive counsel from a physical human therapist.

Additionally, the immersion can lead to feelings of embodiment in VR that can further socially isolate patients. As graphic designers become better at reducing uncanny valley effects in VR, these may be accompanied by a distortion of reality. Patients may come to depend excessively on the VR programme and not want to do these activities in their functional world. After all, performing activities and monitoring them in VR quantitatively is limited to the number of sensors that can be fitted on the patient. Self-reports and other subjective measures can be biased without the presence of a white coat figure.

The implementation of VR in remote patient monitoring is an exciting development and one that should be carefully thought about prior to widespread adoption.

John Christy Johnson is an M.Sc. biomedical engineering candidate at the University of Alberta Rehabilitation Robotics Lab with a background in physiology and neuroscience. His thesis project incorporates elements of telerehabilitation, virtual reality, and wheelchair biomechanics.

If you enjoyed this story, consider reading more below:

Originally published at https://blogs.ibo.org on August 20, 2020.