Letters

New edicts for letters

Restrictions should not be imposed on post-publication peer review

Post-publication peer review should have its place

Brief letters, more letters?

BMJ ought to lead its contributors by example

BMJ may lose correspondents

Restrictions should not be imposed on post-publication peer review

EDITORI often read the short letters in newspapers and journals in preference to the long ones. They are simpler to assimilate, quicker to read, and often amusing too. But I know that the important stuff is generally in the longer letters.

Research published in scientific journals should be open to comment and correction in published correspondence.¹ The BMJ has reduced its word limit for letters from 400 to 300 words (barely more than the length of the abstract of the paper one might be writing about), at the same time as Lancet reduced the time window from eight weeks to two. ^{2 3}

What nexta maximum of 800 words for research papers? More people would read them, I'm sure.

There is an effective "statute of limitations" in leading journals, whereby authors of papers are immune to disclosure of methodological weaknesses once some arbitrary (short) period has elapsed.⁴ Such time limits (four weeks at the BMJ ) discourage post-publication peer review. Similarly, one often cannot adequately address multiple and complex concerns about a study in a short letter.

Convenience should not take precedence over science. Brevity is not more important than scientific rigour. Editors are hiding behind arbitrary rules to avoid having to decide what is actually important. Such restrictions may be suitable for a magazine but not for a scientific journal.

I applaud the rapid responses on bmj.com. It is shocking that so few journals have followed suit. But, although it remains the case that only those published in the paper version get linked to the original article on PubMed, publication in the paper journal should not be determined by rules.

By all means encourage brevity and rapid submissions, but do not rule out longer or later publication if the circumstances warrant it. In particular, letters that draw attention to methodological flaws should not be time limited.⁴

Douglas G Altman, professor of statistics in medicine. 
Cancer Research UK, NHS Centre for Statistics in Medicine, Institute of Health Sciences, Oxford OX3 7LF doug.altman{at}cancer.org.uk

Competing interests: None declared.

Post-publication peer review should have its place

EDITORThe BMJ recently announced a change in its editorial policy concerning correspondence.¹ In an effort to make this section more readable and interesting, letters are no longer be accepted "for their detailed critique of science."

I think that letters discussing the scientific aspects of original articles have always been relatively infrequent and need more encouragement rather than less. ^{2 3} Despite improvements in the assessment of submitted manuscripts, post-publication peer review remains important (for example, identifying statistical errors⁴ and unconsidered potential biases⁵).

The print publication of critical letters is, at present, more likely to correct the medical literatureprinted but not electronic comments are linked to article citations in databases such as PubMed.

Furthermore, in this age of evidence based medicine, when all doctors need to be able to appreciate possible weaknesses of studies, we need to see examples. As readers with limited time may focus on printed letters rather than unselected rapid responses, these examples belong in the print edition.

I suggest fostering more readable and interesting scientific critique rather than simply giving up on it. There must be room for the readers' peer review of published articles.

Sean Bydder, research fellow. 
Sir Charles Gairdner Hospital, Perth, WA 6009, Australia sean.bydder{at}health.wa.gov.au

Competing interests: None declared.

Brief letters, more letters?

EDITORWith reference to Davies's editorial,¹ I have had more than 20 letters published in the Times and can speak with authority on the issue of brevity.

Success also requires speed, but even my taut missives are pared by the correspondence editors. To pack yet more letters into the printed BMJ you must be editorsand edit!

Andrew N Bamji, consultant rheumatologist. 
Queen Mary's Hospital, Sidcup DA14 6LT andrewbamji{at}lineone.net

Competing interests: None declared.

BMJ ought to lead its contributors by example

EDITORDavies says that the new guidelines for letters to the BMJ will still allow the author's voice to remain audible.¹ Many years ago, when editor of the British Journal of Industrial Medicine (as it was then), I found that concept a timely brake on my enthusiasm to "correct the English" in accepted papers.

Davies also encouraged letter writers to be both clear and concise. Yet in the very next issue under "This week in the BMJ" appears this awful specimen:

"The team's appraisal and application of published evidence was crucially informed by a detailed qualitative study of the experiences of students and staff on their course by their own experience as students on other online courses and by informed discussion with other course developers."

Ought not the BMJ to lead its contributors by example?

W R Lee, retired emeritus professor. 
6 Sussex Avenue, Didsbury, Manchester M20 6AQ

Competing interests: None declared.

BMJ may lose correspondents

EDITORWell! Jesus may (or more likely may not) want me for a sunbeam but the BMJ most certainly does not want me for a correspondent.¹

As you are so enchanted with electronics and apparently so disillusioned with the printed word (take note also of the paucity of your book reviews), may I suggest you discontinue the paper version altogether and put me out of my misery.

Robert Richardson, medical author. 
East Pallant Cottage, East Pallant, Chichester, West Sussex PO19 1TZ

Competing interests: RR has no email address, is not connected to the internet, and wouldn't know a website if he was dumped in one. He enjoys reading the printed word and does all his research through this medium.