Human-AI Collaboration in Healthcare: A Scoping Review

Area of Science:

• Healthcare Informatics
• Clinical Decision Support Systems
• Human-Computer Interaction

Background:

• Artificial intelligence (AI) integration into clinical pathways is rapidly increasing.
• Effective human-AI collaboration (HAIC) is crucial for healthcare policy and practice.
• Scoping reviews are vital for understanding the landscape of emerging research areas.

Purpose of the Study:

• To conduct a comprehensive scoping review of empirical studies on HAIC in healthcare.
• To identify trends, evidence gaps, and key factors influencing HAIC effectiveness.
• To inform future research and policy regarding AI in clinical settings.

Main Methods:

• Systematic search of empirical studies on HAIC in healthcare from January 2015 to October 2025.
• Utilized Joanna Briggs Institute methodology and PRISMA-ScR reporting guidelines.
• Included 140 studies from an initial 17,463 identified records.

Main Results:

• Evidence predominantly focuses on diagnostic interpretation, with limited studies in screening, triage, therapeutics, and administration.
• Effectiveness metrics were inconsistently defined and often short-term, task-level.
• Most studies reported benefits for human-AI teams, contingent on task fit, workflow integration, training, and trust calibration.
• Ethical and governance issues were discussed but rarely empirically evaluated.

Conclusions:

• HAIC demonstrates potential benefits, particularly in diagnostics, but requires optimized integration and trust.
• Current research often lacks patient or system-level outcome evaluation and empirical assessment of ethical considerations.
• Future research should prioritize longitudinal, task-specific, and governance-aware evaluations of HAIC in healthcare.