Assessing Large Language Models for Early Article Identification in Otolaryngology—Head and Neck Surgery Systematic Reviews

Key Points

• Outputs from large language models showed inaccuracies, yet some relevant articles were identified.
• The evaluation indicated that recent articles were often overlooked in conventional peer-reviewed methodologies.
• Systematic reviews employing a PRISMA framework serve as the gold standard for literature evaluation.
• Refining large language models for this purpose may improve accuracy and efficiency in literature reviews.