Transparent Reporting of AI in Systematic Literature Reviews: Development of the PRISMA-trAIce Checklist

Area of Science:

• Evidence Synthesis
• Artificial Intelligence
• Research Methodology

Background:

• Systematic literature reviews (SLRs) are resource-intensive, and while AI can accelerate them, it poses transparency challenges.
• Existing guidelines like PRISMA-AI focus on AI as a research subject, not a tool within the review process.
• A need exists for standards ensuring transparency when AI is used methodologically in evidence synthesis.

Purpose of the Study:

• To develop and propose a discipline-agnostic checklist extension for the PRISMA 2020 statement.
• To ensure transparent reporting of AI used as a methodological tool in evidence synthesis.
• To foster trust in AI-assisted systematic reviews.

Main Methods:

• A systematic process was employed to develop the PRISMA-trAIce checklist.
• Literature search for consensus-based AI reporting guidelines (e.g., CONSORT-AI, TRIPOD-AI).
• Extraction, analysis, and thematic synthesis of relevant items to create a modular checklist integrated with PRISMA 2020.

Main Results:

• The PRISMA-trAIce checklist was developed to document AI use in SLRs.
• The checklist covers all SLR sections, including AI tool identification, human-AI interaction, performance evaluation, and limitations.
• It provides specific items for transparent reporting of AI in evidence synthesis.

Conclusions:

• PRISMA-trAIce offers a framework for transparency and integrity in AI-assisted systematic reviews.
• It aims to enhance trust in the responsible application of AI in evidence synthesis.
• The checklist is proposed as a foundation for community consensus and formal endorsement.