Hybrid analysis approach using Graphical Model Structure Learning and Structural Equation Modelling for CloudIoT-based healthcare adoption model in Jordan

This study addresses a gap in understanding and modeling the factors influencing the adoption of CloudIoT-based healthcare (CIoT-H) technology. Despite the potential of CloudIoT technology to enhance healthcare delivery, its utilization remains limited. To address this, we developed a comprehensive theoretical model that examines healthcare professionals' intentions to adopt CIoT-H, considering technological, individual, organizational, and environmental factors. Given the increasing pressures on global healthcare systems, such as aging populations, rising chronic diseases, and shortages of healthcare professionals, this research is timely and critical. We constructed the model using the Analytic Hierarchy Process (AHP) and employed a quantitative methodology to collect data via questionnaires. Data analysis was conducted using Structural Equation Modeling (SEM) and Graphical Model Structure Learning (GMSL), utilizing SmartPLS and BayesiaLab software tools. Initial SEM analysis showed that performance expectancy, effort expectancy, facilitating conditions, perceived privacy, trust, and perceived security significantly influenced healthcare professionals' behavioral intentions. Trust also mediated the effects of performance expectancy, perceived security, and effort expectancy on behavioral intention. Subsequently, GMSL was used to build a data-driven model, which revealed three new relationships that were not considered in the proposed model. These relationships were then incorporated into the model, and the SEM analysis was re-conducted to assess the refined model. The result showed that including these relationships improved the model's goodness-of-fit by decreasing the SRMR value from 0.044 to 0.041. The adjusted R² value for trust increased from 0.799 to 0.842, indicating increased explanatory power. Meanwhile, the explanatory power for performance expectancy, introduced as a new mediator, achieved an adjusted R² of 0.756. Facilitating conditions had the largest effect size on behavioral intention (ƒ2=0.031), perceived privacy on trust (ƒ2=0.28), and effort expectancy on performance expectancy (ƒ2=0.615). Predictive relevance (Q2) was high for all endogenous variables: behavioral intention (0.785), trust (0.837), and performance expectancy (0.755), affirming the GMSL's contribution and the robustness of the refined model. This study contributes to understanding CIoT-H technology adoption by providing valuable insights into the factors influencing its adoption. Furthermore, this study contributes to the body of knowledge by building a theoretical adoption model using the AHP method, which contains factors derived from four main categories. Additionally, it introduces a hybrid approach that combines SEM and GMSL for model analysis and validation. The findings and novelty presented in this research hold significant implications for the domain, guiding policymakers, stakeholders, and healthcare institutions in framing their strategies for implementing and optimizing CIoT-H solutions, ultimately contributing to a more effective and efficient healthcare system.

Text 119964.pdf Download (1MB) Official URL or Download Paper: http://ethesis.upm.edu.my/id/eprint/18495

Actions (login required)

View Item