AI Pipeline for Hospital Quality Improvement Factor Discovery
A new research paper proposes an AI pipeline to formalize hospital Quality Improvement (QI) factor discovery, traditionally a manual, expert-driven process using tools like fishbone diagrams and Lean Healthcare methods. The authors argue that current AI alignment methods assume well-defined tasks, whereas QI factor discovery is exploratory and relies on implicit expert judgments. Their approach treats the task as learning from large language models (LLMs) while preserving exploratory components. The paper is published on arXiv under ID 2604.20055.
Key facts
- Hospital QI translates high-level goals into actionable solutions.
- QI factor discovery identifies key modifiable contributing factors.
- Traditional methods include fishbone diagrams, chart reviews, and Lean Healthcare.
- AI has potential to accelerate QI factor discovery.
- Current AI alignment methods assume well-defined tasks.
- QI factor discovery is exploratory, fuzzy, and iterative.
- The proposed AI pipeline uses LLMs to formalize the process.
- The paper is on arXiv with ID 2604.20055.
Entities
Institutions
- arXiv