3D Telemedicine During COVID-19: Non-clinical Validation

Study Methods The study will involve non-clinical validation with no patients and no interventions. The study will investigate usability, presence, satisfaction and reliability data for 3D telemedicine, and provide subjective feedback in the form of a semi-structured interview.

Participants Healthy volunteers from the staff at Canniesburn Regional Plastic Surgery and Burns Unit, Glasgow Royal Infirmary and West Glasgow ACH, Greater Glasgow and Clyde.

Inclusions and exclusion Healthy staff/students. Recruitment from all contacted, eligible participants will be recorded. Exclusion criteria include visual problems not correctable with glasses/dyslexia.

Consent Participants will be consented in writing for inclusion in the study. The consent will specify that data may be recorded for research and further analysis by the research team, and be available for technical purposes by the research team's industrial partner

Sample Size

30-40 participants for Presence and Usability Questionnaires.

Participants will examine other participants (where one acts as a "doctor" and one acts as a "patient") to provide indicators of usability, presence and task load index. All participants will assess other participants using the screening tasks below (estimated 10 mins maximum):

Screening Tasks: asking "patient" to follow instructions Upper Limb: Functional screening upper limb - hands above head, brush teeth, hands behind back, pro-supination, wrist, prehension grips (power, span, pinch, key, tripod) Lower limb: Functional assessment of lower limb - walk, straight leg raise, take shoes off and examine sole of feet Head: examine worm's eye view of nose, side of nose and top of scalp.

Performing a "task" Perform a form of physiotherapy or equivalent task (chosen at random) such as describing to patient the use of an anterolateral thigh flap for tumour reconstruction.

Outcomes assessed: validated questionnaires, time taken to complete screening tasks, NASA task load index

Randomisation Participants will be randomized to see "patients" in 3D or 2D first. This will be by blocked randomisation using a list and sealed envelopes.

Outcome Measures

Multiple measures of usability and presence will be used. This will be to form baseline data on the 3D telemedicine system (e.g. System usability score >68 is "good"), but also to determine the sensitivity of the various instruments to record overall satisfaction with the systems. This will help to determine the most appropriate instruments for the clinical validation.

Mental Effort Rating Scale - single instrument Likert scale (Paas 1996) University of North Northway - 5 item - telemedicine questionnaire (Buvik 2016) System Usability Scale - 10 item scale - industry standard for usability of technology (Brooke 1986) Presence Questionnaire - 29 item scale - assessment of presence in the system/virtual environment (PQ, Witmer 2005).

Telehealth Usability Questionnaire - 21 item scale - assessment of telemedicine NASA TLX - task load index - 5 item modified scale - an indication of the ease of performing a task and frustration levels.

Overall Telemedicine Satisfaction - VAS Semi Structured exit interview - will discuss with participant their views on the telemedicine system and how we can improve.

Statistics and Analysis

The values obtained for the different rating scales will be summarised, in order to provide a basis for future sample size calculations. With a sample size of 40, the confidence interval around an estimated standard deviation s would be 0.819 s to 1.284 s. With a sample size of 50 it would be 0.835 s to 1.246 (Sim 2012).

The relation between the various rating scales and overall telemedicine satisfaction will be assessed using linear mixed effects regression analysis predicting overall telemedicine satisfaction from the rating scale and type of telemedicine, including a random effect for participant.

The research team will examine whether higher "presence" scores (on Presence Questionnaire) translate to less "frustration" (on NASA TLX), and whether this results in less error/time taken to complete tasks. This will be done using linear mixed effects regression analyses.