A technology acceptance model to predict anesthesiologists' clinical adoption of virtual reality.

ABSTRACT

BACKGROUND: Virtual reality (VR) is a novel tool with demonstrated applications within healthcare; however its integration within clinical practice has been slow. Adoption patterns can be evaluated using a technology acceptance model (TAM). The primary study aim was to use VR TAM to assess factors that influence anesthesiologists' acceptance of VR for preoperative anxiolysis. The secondary aim assessed the model's reliability. METHODS: 109 clinical anesthesiologists at Stanford were exposed to a VR application developed as a distraction tool to reduce preoperative patient anxiety. Anesthesiologists were surveyed about their attitudes, beliefs, and behaviors as predictors of their likelihood to clinically use VR. The primary outcome assessed predictive validity using descriptive statistics, construct validity using confirmatory factor analysis, and standardized estimates of model relationships. The secondary outcome assessed reliability with Cronbach's α and composite reliability. RESULTS: Construct validity and reliability was assessed, where all values established acceptable fit and reliability. Hypothesized predictors of consumer use were evaluated with standardized estimates, looking at perceptions of usefulness, ease of use, and enjoyment in predicting attitudes and intentions toward using and purchasing. Past use and price willing to pay did not predict perceived usefulness. Participants in lower age ranges had higher levels of perceived ease of use than those >55 years. CONCLUSION: All confirmatory factor analysis testing for construct validity had good fit. Perceptions of usefulness and enjoyment predicted an anesthesiologist's attitude toward using and intention to purchase, while perceived ease of use predicted perceived usefulness and enjoyment, attitude toward purchasing and using, and intention to use. Past use and price willing to pay did not influence perceptions of usefulness. Lower age predicted greater perceived ease of use. All scales in the model demonstrated acceptable reliability. With good validity and reliability, the VR-TAM model demonstrated factors predictive of anesthesiologist's intentions to integrate VR into clinical settings.