Page 425 - ISC PROCEEDINGS 21.4

integrative theoretical perspective, the study aims to provide empirical evidence in the
Vietnamese context while offering policy and managerial implications for promoting
effective and sustainable AI adoption in healthcare.
2. Theoretical framework and research methodology
2.1. AI-assisted healthcare
In recent years, the rapid advancement of AI has created substantial potential for
applications in the healthcare sector, contributing to improved efficiency in medical
examination and treatment as well as enhanced experiences for both physicians and
patients (Rajpurkar et al., 2022). AI in healthcare is generally understood as the
application of intelligent algorithms, such as machine learning and deep learning, to
activities including clinical decision support, risk prediction, disease classification, health
management, and treatment optimization (He et al., 2019; Chen, 2024). In this study, AI-
assisted healthcare is defined as the process of healthcare delivery in which AI models
support clinical examination, diagnosis, risk prediction, and prescription optimization,
with AI serving solely as a supportive tool and not as a replacement for physicians.
2.2. Theoretical frameworks
The literature indicates that the PPM framework and the UTAUT framework are
widely employed to explain technology adoption and switching behavior, and therefore
serve as the theoretical foundation for this study. According to Lee (1966) and Moon
(1995), the PPM framework categorizes the determinants of switching into three groups:
push factors (limitations of the current option), pull factors (attractive benefits of
alternative options), and mooring factors (personal and social constraints that hinder
switching decisions). Meanwhile, the UTAUT framework, proposed by Venkatesh et al.
(2003), explains technology adoption through four key constructs, namely performance
expectancy, effort expectancy, social influence, and facilitating conditions, while also
considering moderating variables such as age, gender, and experience.
In the context of healthcare digital transformation in Vietnam, studying the
intention to switch to AI-assisted services requires considering both psychological and
technology-related factors. Therefore, this study combines TPB, UTAUT, and the PPM
model to build a comprehensive framework. The PPM model explains the switching
process: push factors reflect dissatisfaction with traditional healthcare (linked to attitudes
in TPB), pull factors from UTAUT (performance expectancy, effort expectancy, personal
innovativeness) increase intention, while mooring factors (social influence, perceived risk,
switching costs) act as barriers. Overall, PPM provides the structure, TPB explains
psychological aspects, and UTAUT highlights technology perceptions, helping to predict AI
adoption and explain the drivers and barriers of switching behavior.
Therefore, this integrative approach is consistent with current research trends, as
an increasing number of studies advocate for the combination of multiple theoretical
frameworks to simultaneously explain technological, psychological, and contextual factors
(Oliveira et al., 2014).
2.3. Hypothesis development
2.3.1. Push factors
In consumer behavior studies, low satisfaction is viewed as the counterpart to
satisfaction or a state of dissatisfaction, occurring when service performance falls short of
expectations and prompts switching intentions (Oliver, 1980). In the PPM framework, it
acts as a key push factor driving consumers toward alternatives (Bansal et al., 2005; Hsieh
et al., 2012). In healthcare, dissatisfaction with traditional services such as quality, waiting

424