JMIR Publications

ABSTRACT

Background:

The development of AI doctor diagnosis system provides a solution for the allocation of medical resources and the accuracy rate of AI diagnosis has exceeded the level of many primary medical staff. However, many patients are hesitant to accept AI technology. The mechanism of AI doctors' information quality affects patients' adoption intention is still unclear.

Objective:

This study aimed to examine how the information quality of AI doctors influences patients’ adoption intentions. Specifically, drawing on the elaboration likelihood model, this study sought to understand how the diagnostic transparency and argument quality of AI doctors affect patients’ intentions to adopt AI for disease diagnosis and independent decision-making, mediated by perceived expertise and trust.

Methods:

A scenario-based experiment was conducted to investigate the impact of information quality on patients’ adoption intentions. To test the hypotheses, a 2 (diagnostic transparency: low/high) × 2 (diagnostic argument quality: low/high) between-groups experimental design was employed. Each experimental group consisted of 60 valid subjects, yielding a total of 240 valid responses. Data were analyzed using ANOVA and Partial Least Squares.

Results:

Both diagnostic transparency and diagnostic argument quality significantly positively affected patients' perceived expertise. As the central route, the influence of the experimental manipulation of diagnostic argument quality (M1=4.55，M2=5.68，F=58.397，P<.001) on perceived expertise is more significant than that of diagnostic transparency (M1=4.92，M2=5.31，F=5.878，P<.05). At the same time, perceived expertise has a positive impact on trust, trust also positively affected patients’ the intention to adopt AI doctors for disease diagnosis and independent diagnosis, perceived expertise → trust played a full mediating role in the effects of diagnostic transparency and diagnostic argument quality on the two adoption intentions.

Conclusions:

The intention of patients to adopt AI doctor diagnosis systems is pivotal for the efficacy of AI diagnostic outcomes. The findings of this research not only bridge the literature gaps concerning user adoption behavior and algorithm aversion within the medical AI domain but also furnish a decision-making foundation for online intelligent consultation health service platforms, bearing significant practical implications.