Artificial Intelligence as a Safeguard for Clinical Scientific Integrity: A Human-AI Hybrid Model for Medical Peer...

Area of Science:

• Medical publishing
• Scholarly communication
• Clinical evidence integrity

Background:

• Current medical peer review faces challenges like bias, reviewer overload, and potential misconduct.
• These issues compromise the trustworthiness of scientific literature used for clinical decisions.
• Existing peer review systems struggle to maintain high standards of reliability and fairness.

Purpose of the Study:

• To explore the potential of artificial intelligence (AI) and large language models (LLMs) in improving medical peer review.
• To address the limitations and risks associated with AI in scholarly publishing.
• To propose a hybrid human-AI model for enhanced peer review processes.

Main Methods:

• Review of current challenges in medical peer review.
• Analysis of AI and LLM capabilities for manuscript evaluation (e.g., consistency checks, plagiarism detection, compliance monitoring).
• Discussion of AI limitations (hallucinations, lack of judgment, confidentiality risks).

Main Results:

• AI tools can expedite manuscript screening, identify statistical errors, detect plagiarism, and enforce standards.
• Early AI implementations show potential for more thorough, objective, and reproducible reviews.
• AI offers a scalable solution to manage the increasing volume of submissions.

Conclusions:

• AI and LLMs present a significant opportunity to bolster the integrity and efficiency of medical peer review.
• A hybrid model, leveraging AI for routine tasks and human experts for final evaluation, is proposed.
• This human-AI collaboration is crucial for enhancing the quality, fairness, and reliability of the clinical evidence base.