Dear Editor,
We read with great interest the study by Alkhawam et al. [1], that surveyed AI use policies across cardiology and vascular journals.
Inconsistent vascular journal adoption percentages between results and discussion
The results section reports that 113 of 127 vascular journals (89%) had adopted an AI policy, with the remaining 14 journals (11%) lacking one. The discussion later restates these same underlying counts as adoption rates of 78 and 83% for cardiology and vascular journals, respectively, and further describes the non-adoption rates as 22 and 17% for the two journal categories. Since 14 of 127 corresponds to 11% rather than 17%, and 113 of 127 corresponds to 89% rather than 83%, the discussion presents a different pair of figures for vascular journals than the results section derived from the same raw counts, without any indication that a different subset or calculation produced the discrepancy [2].
Self-contradictory presentation of the society-affiliation chi-square result
The society-affiliation analysis is reported as “P = 0.255, P < 0.05, indicating no statistically significant association.” As written, this sentence pairs a specific p-value of 0.255 with the notation “P < 0.05” in the same parenthetical, even though 0.255 does not satisfy that threshold. The stated conclusion that no significant association was found is the correct interpretation of a p-value of 0.255, but the juxtaposition of the reported value with a threshold it does not meet leaves the sentence internally contradictory as it appears in the text, and a reader encountering this line without independently checking the arithmetic could reasonably wonder which figure was intended [3].
Multiple significance tests conducted without a stated multiplicity adjustment
The analysis tests for association between AI policy adoption and at least ten separate variables, including five journal metrics, three access-type categories, quartile ranking, and publication rate, in addition to society affiliation, each of which was evaluated at a conventional significance threshold. No correction for testing multiple associations was described. The quartile-ranking association, reported at p = 0.035, is close enough to the significance threshold that its status could plausibly shift once the number of comparisons performed across the full analysis is considered [4]. However, the discussion presents this finding as a substantive conclusion about higher-ranked journals prioritizing early AI policy adoption without noting this caveat alongside the open-access and hybrid associations reported at more conservative p values.
Against this background, this study offers a genuinely comprehensive census of AI policy adoption across a journal category not previously surveyed at this scale, and its attention to access type and quartile ranking as candidate correlates extends beyond what earlier single-discipline surveys in ophthalmology and radiology have examined. Reconciling the vascular journal percentages between the results and discussion sections, correcting the self-contradictory statistical notation in the society-affiliation results, and noting the number of comparisons tested when interpreting the borderline quartile-ranking association would strengthen confidence in the reported patterns of AI policy adoption.
CRediT authorship contribution statement
S. Ninisha Nels: Conceptualization, Data curation, Formal analysis. Kiruba Wesley: Visualization, Writing – original draft, Writing – review & editing. G.G.S. Pradeep: Visualization, Writing – review & editing, Writing – original draft.
Informed consent
Not applicable.
Ethical approval
Not applicable (Commentary on a previously published article).
Declaration of Generative AI and AI-assisted technologies in the writing process
Generative AI tools were used only for language refinement and formatting assistance. All scientific interpretation, critical appraisal, and manuscript preparation were performed independently by the authors.
Funding
None
Declaration of competing interest
The authors declare that they have no conflicts of interest.
Acknowledgments
None.
Contributor Information
S. Ninisha Nels, Email: sninishanels01@gmail.com.
Kiruba Wesley, Email: kirubawesley@stjosephs.ac.in.
Data availability
No new data were generated or analyzed in this study.
References
- 1.Alkhawam M., Almobayed A., Pandey A., Nanda N.C., Ebrahimi A.J., Ahmed M.I. Exploring AI use policies in manuscript writing in cardiology and vascular journals. Am. Heart J. Plus. 2025;58 doi: 10.1016/j.ahjo.2025.100586. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.de Borst G.J., Boyle J.R., Dick F., Kakkos S.K., Mani K., Mills J.L., Björck M. Publication standards for reporting vascular surgical research. Eur. J. Vasc. Endovasc. Surg. 2024 doi: 10.1016/j.ejvs.2024.10.009. [DOI] [PubMed] [Google Scholar]
- 3.Hunter M.K., Russell A.J., Maldonado G., Burstyn I. Exploring the proper use of p-values and confidence intervals in leading epidemiology journals. Global epidemiology. 2026;11 doi: 10.1016/j.gloepi.2026.100247. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Mrkvička T., Konstantinou K., Kuronen M., Myllymäki M. Global test for covariate significance in quantile regression. Stat. Comput. 2026 doi: 10.1007/s11222-025-10774-9. [DOI] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
No new data were generated or analyzed in this study.