Seite - 70 - in Applied Interdisciplinary Theory in Health Informatics - Knowledge Base for Practitioners

Holden [8] lists several other examples where TAM or UTAUT were extended by context-specific variables. This shows that despite their large popularity, both models may need to be parsimoniously applied in more complex health care settings. Holden also points to the fact the key variables of TAM and UTAUT are not measured uniformly in different studies. Instead, studies often modify original survey items to adapt the questions to the local study context (either by rewording questions or by adding completely new questions). All this shows TAM and UTAUT are somewhat unspecific for health care settings. In general, both TAM und UTAUT have been found to predict Intention to Use quite well, with explained variance up to 70%. Yet, closer analysis to their application in health care by Holden (2010) shows that only Perceived Usefulness was consistently found to be a significant predictor of Intention to Use (in all of the 16 reviewed studies) [8]. In contrast to this, Perceived Ease of Use was found to be a significant predictor of Intention to Use in seven of 13 studies only [8]. And Social Influence, an UTAUT variable, was found to be significant predictor in four of eight studies only. Also Gücin (2015) states, based on a literature review, that Perceived Usefulness is “the most powerful predictor of the technology acceptance” [15]. Summarizing these findings, we see that the key assumptions of TAM and UTAUT could not be confirmed in a large number of technology acceptance studies in health care. These findings indicate that health care is indeed a special setting where the simple assumption of TAM and UTAUT may not fully match the more complex reality. Holden (2010), for example, summarizes that Perceived Ease of Use may not be that important for technology acceptance and usage when users are sufficiently experienced with the system or when they have sufficient IT support. Also, Social Influence may not influence physicians as users so strongly, as they are more independent and “immune to peer pressure” [8]. Also, after an analysis of several acceptance theories, Gücin (2015) points to the fact that the acceptance factors for health care professionals and patients may be different, with patients seeing for example ease of use as more important than health care professionals [15]. Also, he argues that important acceptance factors such as suspicions of confidentiality and privacy or individual characteristics of the user (e.g. of early adopters) may be strong influencing factors, but these are not considered in the original models [15]. To conclude, while TAM and UTAUT have been broadly adopted as a means of predicting technology acceptance and usage, the findings in health care are quite mixed. Both the fact that many studies in health care cannot find support for some basic hypothesis of TAM and UTAUT, and the fact that many authors added variables to the original TAM and UTAUT models or revised the survey instruments to respond to context influencing factors, point to the fact that the original TAM and UTAUT fail to demonstrate strong predictive capabilities for technology acceptance in health care [14]. 4. Conclusion TAM and partly UTAUT provide a more technology-centered view on technology acceptance, where acceptance is understood to mostly depend on the nature of technology [14], i.e. functionality and ease of use. Socio-organizational, workflow, cultural or emotional aspects as well as differences in user groups (physicians, nurses, patents) are not well covered [14] and may explain why in several studies in health care, basic assumptions of the model could not be confirmed. E.Ammenwerth /TechnologyAcceptanceModels inHealth Informatics: TAMandUTAUT70