L-PRISMA: An Extension of PRISMA in the Era of Generative Artificial Intelligence (GenAI)
arXiv:2603.19236v1 Announce Type: cross Abstract: The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) framework provides a rigorous foundation for evidence synthesis, yet the manual processes of data extraction and literature screening remain time-consuming and restrictive. Recent advances in Generative Artificial Intelligence (GenAI), particularly large language models (LLMs), offer opportunities to automate and scale these tasks, thereby improving time and efficiency. However, reproducibility, transparency, and auditability, the core PRISMA principles, are being challenged by the inherent non-determinism of LLMs and the risks of hallucination and bias amplification. To address these limitations, this study integrates human-led synthesis with a GenAI-assisted statistical pre-screening step. Human oversight ensures scientific validity and transparency, while the deterministic nature of the statistical layer enhances reproducibility. The proposed approach sy
arXiv:2603.19236v1 Announce Type: cross Abstract: The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) framework provides a rigorous foundation for evidence synthesis, yet the manual processes of data extraction and literature screening remain time-consuming and restrictive. Recent advances in Generative Artificial Intelligence (GenAI), particularly large language models (LLMs), offer opportunities to automate and scale these tasks, thereby improving time and efficiency. However, reproducibility, transparency, and auditability, the core PRISMA principles, are being challenged by the inherent non-determinism of LLMs and the risks of hallucination and bias amplification. To address these limitations, this study integrates human-led synthesis with a GenAI-assisted statistical pre-screening step. Human oversight ensures scientific validity and transparency, while the deterministic nature of the statistical layer enhances reproducibility. The proposed approach systematically enhances PRISMA guidelines, providing a responsible pathway for incorporating GenAI into systematic review workflows.
Executive Summary
This article proposes an extension of the PRISMA framework, leveraging Generative Artificial Intelligence (GenAI) to enhance systematic review workflows. The proposed approach, L-PRISMA, integrates human-led synthesis with a GenAI-assisted statistical pre-screening step, addressing limitations in reproducibility, transparency, and auditability. By combining human oversight with a deterministic statistical layer, L-PRISMA provides a responsible pathway for incorporating GenAI into systematic review workflows, ultimately improving efficiency and time while maintaining scientific validity and transparency.
Key Points
- ▸ The article proposes a new framework, L-PRISMA, which extends the PRISMA guidelines to incorporate GenAI.
- ▸ L-PRISMA integrates human-led synthesis with a GenAI-assisted statistical pre-screening step.
- ▸ The approach addresses limitations in reproducibility, transparency, and auditability, while maintaining scientific validity and transparency.
Merits
Strength in Addressing Reproducibility Challenges
The article effectively acknowledges and addresses the limitations of GenAI in systematic reviews, particularly in terms of reproducibility, transparency, and auditability.
Enhanced Efficiency and Time
The proposed approach has the potential to significantly improve efficiency and time in systematic review workflows, making it more appealing to researchers and practitioners.
Human Oversight and Transparency
The integration of human-led synthesis ensures scientific validity and transparency, which is essential for maintaining the credibility of systematic reviews.
Demerits
Potential Bias and Hallucination Risks
The article highlights the potential risks of bias amplification and hallucination associated with GenAI, which may compromise the validity of systematic reviews.
Technical Complexity and Implementation Challenges
The proposed approach may be technically complex and challenging to implement, particularly for researchers without expertise in GenAI and statistical modeling.
Evaluating the Effectiveness of L-PRISMA
The article does not provide empirical evidence to support the effectiveness of L-PRISMA, which may limit its adoption and implementation.
Expert Commentary
While the article makes a compelling case for the potential benefits of L-PRISMA, it is essential to acknowledge the limitations and challenges associated with incorporating GenAI into systematic review workflows. The proposed approach requires careful evaluation and validation to ensure that it maintains the scientific validity and transparency of systematic reviews. Furthermore, policymakers and researchers must address the broader implications of GenAI on the research landscape, including the potential risks of bias amplification and hallucination.
Recommendations
- ✓ Empirical evaluation of L-PRISMA to assess its effectiveness and limitations.
- ✓ Development of guidelines and standards for the use of GenAI in systematic reviews, including protocols for addressing reproducibility, transparency, and auditability.
Sources
Original: arXiv - cs.AI
