User acceptance of telerehabilitation in Germany: a structural equation modeling approach based on the UTAUT2 model

The use of telerehabilitation in the field of medical rehabilitation is increasing, particularly following the coronavirus pandemic, which accelerated the digitization of rehabilitation services. Patient acceptance is crucial for the sustainable implementation of these digital services. While most studies in the rehabilitation context focus on behavioral intention (attitudinal acceptance), this study additionally examines factors influencing use behavior (behavioral acceptance) in a real-world care setting. It also investigated patients' usability experience and affinity for technology.

A cross-sectional design was applied, recruiting participants from thirteen rehabilitation centers. Patients could participate in the survey digitally or via paper. Acceptance was measured based on the UTAUT2 model, supplemented with questions on privacy concerns, usability, and technology affinity. Data was first descriptively analyzed, followed by structural equation modeling to determine factors influencing behavioral intention and use behavior.

The analysis included 230 cases. Participants' affinity for technology was in the medium range (M = 3.45, SD = 1.17; 1 = strongly disagree, 6 = strongly agree), while acceptance of telerehabilitation could be classified as high (M = 2.37, SD = 1.36; 1 = strongly agree, 7 = strongly disagree). The evaluation of the UTAUT2 model showed that the constructs performance expectancy (p < 0.001), habit (p < 0.001) and hedonic motivation (p < 0.001) had a significant positive impact on behavioral intention, while habit was the only construct that significantly influenced use behavior (p = 0.031). The model showed good explanatory power (R ² = 0.746) for behavioral intention but limited explanatory power for use behavior (R ² = 0.049). The constructs habit (f²  = 0.117), hedonic motivation (f²  = 0.085) and performance expectancy (f²  = 0.120) contributed significantly to the predictive quality of the model.

Only a few constructs of the UTAUT2 model, namely performance expectancy, hedonic motivation and habit, significantly influenced behavioral intention in the German telerehabilitation context. Our results also showed that the hypothesized relationship between behavioral intention and use behavior was not statistically supported. This finding may reflect the "intention-behavior gap" described in the literature, in which behavior intentions do not necessarily lead to actual use.