All journals should include a correspondence section

Abstract

Letters to the editor can provide useful scientific information and evaluation of published work as well as acting as an additional level of peer review. Furthermore, letters are good reading material, especially if they involve a debate between authors. Finally, letters are relatively short. Therefore, inexperienced career researchers can use such an opportunity to practice putting together a cogent argument. However, it is far from an ideal situation if letters are the only (or main) type of article on which to base an academic career.

Key Words: Correspondence, Journals, Letters to the editor, Medical writing, Peer review, Debate

Core tip: Letters provide another level of worldwide peer review. Three editors express their opinions regarding the scientific value and structure of correspondence sections in journals.

INTRODUCTION

In this brief overview, three editors express their opinions regarding the scientific value and structure of correspondence sections in journals. Interpretations and suggestions are based on experience and the literature.

THE NEED FOR A CORRESPONDENCE SECTION

We propose that a correspondence section is an essential part of all journals. The reasons are summarised as follows[1-3]: Letters provide an additional level in the peer review process. Essentially, anyone worldwide can comment on a publication. Letters often promote good reading, especially when they involve a debate between authors. This is especially true for journals that have letters openly available. Given that letters are short, they are relatively easy to write. Therefore, they provide a training opportunity for inexperienced authors. Letters do not count as items when the Clarivate journal impact factor is calculated but if they are cited, these citations count. Thus, any citations of a letter may prove helpful for journals. However, we also need to consider that most letters are probably not highly cited.

TIPS ON WRITING A LETTER

A general rule would be a short text (the shorter, the better); brevity is important[1,2]. Therefore, letters need to focus on a restricted number of topics. Most journals impose limits on the word count and number of references. Some journals allow inclusion of a figure or table in a letter[1,2]. However, some editors provide substantial flexibility. Most letters are usually related to publications in the same journal[1-3]. Indeed, some editors do not consider letters unless they relate to material published in their journal. There are broadly two types of letters[1].

Correspondence

This is the commonest type. Such letters aim at one of the following goals[1-3]: (1) To contradict a published finding, for example by citing omitted studies or presenting unpublished results. Letter authors may also wish to highlight methodological or statistical flaws in a published study; (2) To reinterpret a published finding; for example based on additional findings; and (3) To support a published finding; for example on the basis of additional findings, possibly unpublished. This may include indirect evidence (e.g., involving a different gender, ethnicity, species, methodology or related disease).

Early unpublished findings or a case report/series

More rarely, letters present early (unpublished) findings or a case report/series[1,2]. Such letters are miniatures of full studies or case reports. Their main advantages for the authors include rapid publication and the ability to present data on smaller patient series[1,2]. Full papers take longer to be published and processed. This may even take several months and it is possible that during that time more recent and relevant findings become available. For the journals, a potential advantage of full papers (and reviews) is that they are likely to have a higher citation rate than letters.

One final tip for academics and clinicians: avoid exclusively writing letters to the editor without also authoring original or review articles[4]. Indeed, it has already been noted that some authors try to build their career solely on letters published in high-ranking journals[4]. This will be noticed by others and will not be to the authors’ benefit.

SUGGESTIONS FOR JOURNAL EDITORS REGARDING MANAGING A CORRESPONDENCE SECTION

We suggest that all journals could benefit from a correspondence section as a peer review “safety net”. One of us has resigned as Associate Editor from two journals, because they would not introduce a correspondence section on the grounds that it would require too much editorial work.

A dedicated editor for the correspondence section would be ideal. However, this may be impractical for some journals. One of us has recently experienced a 5-mo delay regarding a decision on a 300-word letter. In our opinion, this represents completely unacceptable standards by the editorial staff of this journal. However, this is probably and hopefully, a rare event.

Letters provide an opportunity for a rapid response by journal editors[1,2]. Based on our experience both as editors and authors, this may be, at least ideally, a matter of a few days. When letters refer to a specific publication, the authors of the latter usually provide a response, pointing out every possible error.

What to do if authors decline to respond to a letter commenting on their work? There is no simple answer. Possibly, if a letter is highly critical of a study, it may be published together with an editorial message, stating that the authors of the original work declined to respond. It would be unfortunate if some authors avoid criticism just by refusing to respond to valid points raised in a letter. Again, this has happened to us, although the definition of valid comments is based on our knowledge/views. Nevertheless, in our opinion they were obvious. That is why, in similar circumstances, we prefer to underline that the letter containing criticisms will be published, whether the authors of the original work respond or not. Editors must not suppress valid criticism of a publication thinking that it may suggest an oversight of errors by the peer reviewers and editors involved. This is an example of how correspondence provides another valuable level of peer review. One of us is currently involved in resolving such a problem. Obviously, any improvements in peer reviewing are welcome, and are still being sought[5-7].

In defence of authors who refuse to respond to comments in a letter, we need to consider that responding may require considerable additional work, which they do not wish to carry out or would like to reserve for their next publication. In such circumstances, honesty is the best policy. The authors can just state why they cannot provide a detailed response at this time, but they will do so in their forthcoming work. However, the comments will remain in the literature. If they are not covered by future work, this deficiency may be pointed out. Citing an older letter to show that the queries raised were answered is not only professional behaviour, but will also suit the journal where the letter was published by delivering a citation.

Other editorial issues include whether to allow more than one round of exchanges regarding the same publication. The time allowed between publication of an item and the submission of related letters needs to be clearly stated in the instructions for authors.

Finally, in the event of an interesting but too long letter, an option may be to convert it to a commentary or brief communication.

CONCLUSION

Letters to the editor are useful for authors, readers and journals. They provide training for younger researchers and are another valuable level of peer review. For all these reasons, in our opinion as editors, a correspondence section is likely to be a useful part of all scientific journals.