In the more than two-century-old history of medical publishing, there has probably never been a time when the role of the journal editor has been as important as it is today. The onslaught of artificial intelligence (AI) has added to preexisting challenges, and in a metrics-driven publishing industry, editors are burdened with concerns beyond the simple pursuit of scientific truth. Nevertheless, given that nearly 3.3 million scientific articles were published globally in 2022, of which a substantial fraction pertained to the health sciences, we consider it prudent to highlight some issues that editors face in performing their duties.1 In 2022, India contributed more than 200,000 published scientific articles, only surpassed by the United States and China.1 However, there have been questions raised widely about the quality of research output as opposed to the volume, with some arguing that Indian scientific journals lag due to the generally lower quality of overall research, and the best Indian research gets published in leading international journals.2 This places an additional onus on all of us to raise the bar of our work in the national interest, if not for other reasons. One common issue encountered when managing articles from the Global South for the JIAG and other journals is deviation from international standards of writing manuscripts. Nonadherence to guidelines such as STROBE and CONSORT makes papers clumsy to read. Often, new authors resort to writing substantial parts of papers in bullet points or unfinished sentences, instead of adopting the traditional style of writing in paragraphs. While this is satisfactory for synopses of theses, project submissions for funding, and ethical committee submissions, this gives a shabby look to articles published in journals. The terminology and writing style used in the clinical discussion section of case reports and case series differ significantly from those used in bedside rounds, but we regularly find authors unaware of this. Writing a scientific article is learned on the job, and with intensive practice, and the best way to get on with it is going through high-quality papers in leading international journals, preferably those from scientific societies or with high impact factors, and absorbing the nuances of presenting data in acceptable forms. Good advice for young researchers is to read more before writing. The importance of the EQUATOR network and its extensive guidelines cannot be overstated, and adhering to the provided checklists can help produce a high-quality, publication-ready article.3 Another prudent idea may be to involve a statistician at all phases of research, from inception to writing the article, to avoid uninterpretable results, improper use of P values, and lack of justification of sample size. Once an article is returned from peer review with concerns about some of the statistics, it is better to review the analysis in its entirety before resubmission, rather than addressing only the reviewer’s concerns. As journals rise in stature with indexing, a natural increase in submissions is expected. Often, busy editors have to make quick judgments for and against sending an article out for peer review. They have to rely on the cover letter and abstract – two immensely neglected parts of scientific manuscripts. The cover letter and abstract are like “an elevator pitch,” and due diligence in framing these by highlighting the merits of the manuscript increases the chances of peer review and eventual acceptance. It is often disappointing to see misstated journal names, typographical errors, and a general casualness in the rebuttal letter to reviewers. We have had cases of authors using AI to generate rebuttal letters – a highly discouraged practice and easily picked up by regular editors and peer reviewers. Put simply, it increases the likelihood of an outright rejection of the article. One area where AI-based grammar checking tools can be put to good use is correcting English language syntax. Machine-generated language sounds mechanical to the editors’ and readers’ ears, and generative AI is again discouraged in framing entire sections of manuscripts. Using AI as an assistant is prudent, rather than as a guide as exemplified in Figure 1. The JIAG, too, is poised to develop an AI-use policy for the future.Figure 1: A relaxed editorial mind. An editor is truly relaxed only when authors and reviewers adhere to the points mentioned in the boxes: the two boxes at the bottom pertain to reviewers and the others to authors Figure prepared with assistance from an AI-based image generation tool. Design idea and all written textual points are provided by the author; AI only helped generate the image without any technical contributionWith these basics covered, comes the important question of what attracts reviewers and editors of a geriatric-centric journal? “Hot topics” such as antiamyloid therapy in Alzheimer’s disease and AI in healthcare obviously do; so do less explored study techniques such as qualitative research, mixed-methods analysis, and randomized trials of hitherto unexplored therapies in the context of the particular country (India and the Global South for the JIAG). Small sample sizes often lead to rejection when the study is simply a duplication of prior research, but they may still be relevant for novel topics. A new technique being adopted by editors of some journals is to reinvite manuscripts with some of these deficiencies as letters to the editor instead of full articles. The editors and authors are two pillars of the publication process. The third very important pillar is the reviewers. And, this is where modern publishing faces its greatest challenge, because everyone wants to publish, but no one wants to review. While some authors have been of the opinion that peer review is a thankless job, it has to be remembered that Web of Science gives options to add successfully performed peer reviews to one’s research profile, and this does bolster the credentials of those willing to pursue medical academics. The advent of AI has again hit the peer review process hard. Reviewers’ services are sought from experts of the field to contribute to the improvement of submitted articles. If it was solely about ticking the review checkbox, the entire task of the AI-based review process could be handled by the publisher itself, with no need for even editors. The expectation from a peer reviewer is not to do a proofreading of the article with a fine-toothed comb, but rather to provide critical insights and point out obvious flaws in study design or data presented. As one of the letters by Surya Kant Tiwari in the current issue of the JIAG discusses, AI suffers from the complex problem of hallucinations or generating false data, which may appear true on cursory inspection – one more reason to abstain from AI-generated reviews. Uploading of manuscript data to AI tools may also lead to ethical and confidentiality concerns. Many prominent journals initially operated completely driven by editorial teams, and a switch to mandatory peer review occurred only with rising volumes of submissions.4 Peer reviewers are, thus, an extension of the editorial team and hold great power as custodians of high-quality science. In this editorial, we addressed editorial expectations from authors and peer reviewers and briefly discussed the issues posed by AI. As we venture together into the rapidly changing world of scientific publishing, newbies and old hands of the process both need to rise to meet the challenge. We leave you with a glimpse of the relaxed mind of a journal editor Figure 1.
Mathur et al. (Thu,) studied this question.