Title : A mixed-methods study on the current state of information overload and its influencing factors among patients with stroke
Abstract:
Discrete choice experiments (DCEs) have increasingly been used to quantify preferences for artificial intelligence (AI)-based healthcare services. However, evidence on how DCEs have been designed and applied in this field remains fragmented.
Methods: Systematic searches were conducted in Chinese and English databases (e.g., CNKI, PubMed, Web of Science) covering the period from the databases’ inception to March 2026. Original studies published in Chinese or English that used DCEs to investigate preferences for AI-based healthcare services were included. Through literature screening, data extraction, and quality assessment, the study subjects, attribute and level definitions, data analysis models, and primary outcomes were systematically summarized.
Results: Twenty-three studies were included. Study populations included patients, the general public, and healthcare professionals. Application scenarios covered cancer screening, imaging diagnosis, chronic disease management, mental health, remote health management, and intelligent consultation. Attributes were grouped into 12 dimensions, with technical performance, human–AI collaboration, and economic burden being most frequently examined. Conditional logit, mixed logit, and latent class models were commonly used. Across studies, respondents generally valued accuracy, reliability, affordability, and physician involvement. Most patients and members of the public preferred AI–physician collaborative models over fully autonomous AI services.
Conclusion: DCEs provide a useful approach for quantifying multi-attribute preferences for AI-based healthcare services. Future studies should improve attribute development, experimental design, and reporting transparency, and should pay greater attention to privacy, fairness, accountability, and interpretability.
Keywords: Artificial Intelligence; Discrete Choice Experiment; Healthcare Services; Choice Preference; Scope Review
