JMIR Publications

ABSTRACT

Background:

Background AI-driven tools have shown promise in assisting patients with healthcare navigation and improving e-health literacy, but research systematically examining their acceptance from the perspective of patients remains limited.

Objective:

Objective Assess patients’ barriers to healthcare navigation, e-health literacy, and willingness to use AI-assisted nursing services, and use a mixed-methods approach to develop evidence-based nursing policy recommendations.

Methods:

Methods The study employed an explanatory mixed-methods design. The first phase consisted of quantitative analysis, using hierarchical multiple regression to identify independent predictors of use intention. In the second phase, purposeful sampling was conducted based on the results of the first phase, followed by semi-structured interviews, which were analyzed using Braun and Clarke’s thematic analysis method. Data were integrated using connecting and merging strategies, and the results were presented in a joint presentation matrix.

Results:

Results Perceived usefulness and e-health literacy were positive predictors of AI use intention, while perceived risk was a negative predictor. Barriers to accessing medical care did not reach statistical significance in the final model. The qualitative analysis identified five dimensions and 15 subthemes.

Conclusions:

Conclusion Perceived usefulness and e-health literacy are important factors driving patient’s use intention of AI-based healthcare services, while multidimensional perceived risks—including distrust of technology, privacy concerns, and emotional detachment—constitute the primary barriers. The digital divide places older adults with chronic conditions, who have the most urgent healthcare needs, at a disadvantage when it comes to AI applications. Healthcare policies should promote a service model that combines AI systems with human triage and establish a nurse-led process for reviewing AI outputs. Clinical Trial: Not applicable