Objective:

To report the prevalence of fabricated references in biomedical literature and assess their implications for evidence integrity, particularly in clinical decision-making.

Key Findings:

• 4,046 fabricated references found across 2,810 papers.
• Fabricated references increased from 1 in 2,828 papers in 2023 to 1 in 458 papers in 2025.
• Quarterly fabrication rate rose from about 4 to nearly 57 per 10,000 papers.
• Review articles had a fabrication rate 57% higher than other publication types.
• 98.4% of affected papers had received no publisher action.

Interpretation:

The rise in fabricated references poses significant concerns for the integrity of medical literature, potentially undermining clinical decision-making by introducing fictitious evidence.

Limitations:

• Verification pipeline evaluated only references with PubMed identifiers, excluding 23% of structured references, which may impact overall findings.
• Estimates reflect precision rather than recall, meaning undetected fabricated references were not counted.
• Early 2026 data is preliminary and limited to a short time frame.

Conclusion:

Automated reference verification tools, such as [specific tools], could enhance the integrity of medical publishing, and retrospective screening of existing literature is recommended.