Risk-Tiered Governance for Hospital Artificial Intelligence: A Framework Synthesis and Implementation Pathway

Area of Science:

• Health Informatics
• Artificial Intelligence in Healthcare
• Clinical Informatics

Background:

• Artificial intelligence (AI) is increasingly integrated into various hospital workflows, including EHRs, imaging, and operational management.
• Effective governance is crucial for managing the risks associated with hospital AI applications.

Purpose of the Study:

• To propose a risk-tiered governance framework and implementation pathway for hospital AI applications.
• To provide an adaptable scaffold for prioritizing AI oversight based on risk, resources, and organizational maturity.

Main Methods:

• A narrative review and framework synthesis of peer-reviewed evidence, guidelines, and case reports.
• Utilized systematic reviews for risk identification, empirical studies for implementation patterns, and case reports for organizational design.

Main Results:

• A four-component framework: use-case inventory, six-domain risk taxonomy, four-tier risk scheme, and governance architecture.
• Proposed a patient safety-oriented lifecycle pathway from initiation to retirement and organizational learning.
• Included barriers, first steps, and a maturity-based ramp-up model for adoption.

Conclusions:

• Hospital AI governance should be proportional lifecycle control, not a one-time approval.
• The framework is adaptable, intended to guide leaders in prioritizing AI oversight.
• Emphasizes managing AI risks based on potential harm, automation, reversibility, and operational coupling.