The dual path of the technology acceptance model: An application of machine learning cardiotocography in delivery rooms

Cardiotocography (CTG) is widely used for assessing fetal well-being, with the aim to reduce the risk of unnecessary and invasive obstetric interventions and to improve maternal-fetal outcomes. Artificial intelligence technologies could improve the efficacy of CTG; however, the acceptance of such technologies by clinicians is a prerequisite for their sustained adoption in clinical practice. In this chapter, we present a study aiming to investigate the factors associated with the acceptance of an AI-based tool aimed at interpreting the CTG by a group of midwives. An online survey was conducted with 302 midwives. The mean age of the respondents was 34.21 years (SD=9.71), and they reported an average practice of 10.08 years (DS=9.86). The questionnaire included some hypothesized predictors of acceptance, such as personal responsibility, self-efficacy at work, perceived utility, and easiness of use. The results showed that perceived usefulness and perceived easiness of use showed positive and significant effects on the intention to use the AI technology. Moreover, in the final model, personal responsibility showed a higher effect on perceived utility, while self-efficacy showed a higher effect on the easiness of use. These results confirmed some previous models of technology acceptance also for an AI-based tool aimed at interpreting the CTG. Moreover, it adds to the previous literature, emphasizing the role of clinicians’ perceived responsibility and self-efficacy in explaining the technology acceptance process.