Human-AI Collaboration in Healthcare: Navigating Trust, Autonomy, and Clinical Decision Support
Gunjan Kumar, Swati Pandey, Shatrudhan Prajapati, Ajay Pal Singh, Shikha YadavIntroduction:
Artificial Intelligence (AI) is increasingly shaping healthcare, particularly through Clinical Decision Support Systems (AI-CDSS). These tools help physicians interpret complex data and guide patient management. Despite their promise, adoption remains limited due to concerns about trust, professional autonomy, and usability. Existing studies often examine these issues in isolation, leaving gaps in understanding how they interact.
Methods:
This narrative review applied the Unified Theory of Acceptance and Use of Technology (UTAUT) and the Technology-Organization-Environment (TOE) frameworks to synthesize findings from literature published between 2010 and 2025. Databases searched included Pub- Med, Scopus, IEEE Xplore, SpringerLink, and ScienceDirect. Studies addressing adoption, trust, and usability of AI-CDSS were mapped to UTAUT constructs (e.g., performance expectancy, social influence, technology anxiety) and TOE dimensions (technological, organizational, environmental).
Results:
The review highlights that the successful adoption of AI-CDSS depends on more than technical accuracy. Physicians’ trust is shaped by system transparency, explainability, and perceived reliability. Organizational support, including training and infrastructure, strongly influences readiness. Concerns about autonomy and liability remain central barriers, while personal innovativeness and social influence encourage adoption.
Discussion:
Findings suggest that AI is best positioned as a supportive “co-pilot” rather than a replacement for clinicians. Explainable and user-friendly designs, combined with humancentered interfaces, can mitigate technology anxiety and preserve professional judgment. Ethical oversight, bias mitigation, and data governance remain essential to responsible integration.
Conclusion:
Adoption of AI-CDSS requires a balance of innovation and empathy. By aligning technical performance with trust, autonomy, and institutional readiness, AI can be integrated responsibly into healthcare. Future efforts should focus on interdisciplinary design, continuous training, and patient-centered transparency to ensure equitable and ethical use.