Changes in ethics and transparency reporting in dental journal articles between 2021 and 2025: a full-text...

Explicit reporting of research ethics and transparency practices is increasingly emphasized in biomedical publishing, but whether recent changes have occurred uniformly across reporting domains remains unclear.

We conducted a full-text, cross-journal text-mining study of 240 dental journal articles published in 2021 and 2025, with 20 articles per journal-year stratum across six established dental journals. Full text was screened for eight predefined markers: ethics approval, informed consent, conflict-of-interest disclosure, funding disclosure, data availability, trial registration, animal ethics, and AI/LLM disclosure. Automated detection was checked against single-reader manual coding of a 36-article sample (96.2% agreement). Prevalence was compared using Fisher exact tests. Reported percentages are corpus-level detection rates across all 120 articles per year, not compliance rates among the studies to which each marker applies.

Data availability showed by far the largest change, increasing from 4.2% to 47.5% (p < 0.001) and was the only marker with a statistically robust increase. Ethics approval (53.3% to 69.2%), funding disclosure (58.3% to 73.3%), and animal ethics (1.7% to 8.3%) showed smaller directional increases. Conflict-of-interest disclosure was near-universal in both years (95.8% and 97.5%). Informed consent and trial registration showed little change. AI/LLM disclosure was detected only in 2025 (0% to 4.2%). Mean detected markers per article increased from 2.73 to 3.55, and no 2025 article reported zero markers.

These findings indicate uneven changes in dental reporting, with the clearest gains in data availability statements and little change in ethics approval, consent, or trial registration.