Reviewing the “Letter-to-editor” section in the Bulletin of the World Health Organization, 2000-2004
Letters to the editor 2000-2004
10 May 2006 - Letter section of a scientific journal is an essential part of post- publication peer review. (1) Many readers seem to assume that articles published in peer reviewed journals are scientifically sound, despite much evidence to the contrary. Hence, misleading work should be identified even after publication. (2) Therefore, journals welcome and encourage the submission of letters to editors either a critical or research letters. (1) The aim of this work was to conduct a descriptive retrospective biblometric study where correspondence section over 5 years period starting 1st January 2000 to 31st December 2004 in the Bulletin of the World Health Organization were reviewed and analyzed. A PubMed search on the total number of publications and percentage of letters to editors from it was performed through including the name of the Bulletin and letting it directed to PubMed within a single hour limit (December 11th 2005). The PubMed search was limited between January 1st 2000 to December 31st 2004 to get to the total number of publications in the journal and then further limit the search to “Letter” as a type of publication. The search strategy was as follows: “ Search Bulletin of the world Health Organization [ta] Field: All fields, limits: Publication Date from 2000/01/01 to 2004/12/31, Letter”.
The PubMed search revealed that the Journal published 1163 articles over the studied period, of them 65 were “Letters”. Of the 65 letters, 52 represented critical letters, 12 research letters, and one WHO news. All the 65 letters were reviewed and analyzed after printing it from the journal website.
The letters constituted only 5.6% of the overall publication in the journal over the studied 5 years. About 45.3% of letters were single authored with author number ranged from 1-7 (median =2). Of the first three authors- if there-, only one author (.085%) wrote more than one letter in the Bulletin during the studied period. The number of references in all letters ranged from 0 to 10 with a mean of 4.21 (SD=2.49) and a median of 5. About 31.5% of the letter authors quoted themselves in their letters. The Bulletin published the reply of the original author on a letter criticizing his original article in the same issue in one case. Besides, four letters replying to 3 critical letters and an editorial comments were published in separate issues. For the rest 47 critical letters, there was no evidence of the original author reply in the same or other issues. However, there is a possibility that original authors replied critical letter writers in issues not reviewed after 31st December 2004. The duration between publishing the letter and date of publication of the original study ranged from 1 to 22 months with a mean of 6.85(SD=4.76). The commonest reasons for writing a critical letters were to give one’s own or country experience, add new information, raise new issues, or criticize conclusion reached. Only one letter criticizes statistical analysis or require appropriate one. (Table 1)
The current study highlights important issue as duration for publication, percentage of authors with more than one letter during the studied period and the commonest reasons to write a letter. In Mayberry study (3) the reviewed authors quoted themselves in 44% of letters, whereas only 32% of the authors did in the current study. The majority of letters in the current study criticize an original study where no critical letter totally agreed with the original study. Winker and Fontanarosa 1 in JAMA stated that letters that simply applaud the authors work generally are not published because letters are not a vote on article’s merits. However, a relatively old audit in BMJ of correspondence column in almost 4 months showed that broadly equal numbers of letters agreed and disagreed with articles, irrespective of section.(4) Similarity was noticed between the commonest reasons for writing letters between the current study and Caswell study in the Medical journal of Australia.(5) The noteworthy finding in the current study is that only one of the reviewed letters scrutinized statistical analysis or request appropriate one in the original study.
Finally, I wish to discuss the delay between publication of articles and their critical letters. Such delay interferes with the educative importance of the letters, which in turn alters the significance of the article to non-expert readers. Even 5 months could be considered a delay by some authors.(6) To overcome the delay in letter publications international journals set either time limitations for submission, word limitation for processing and number of letters accepted or limitation to the way of submission. (7) Maybe launching e-letters’ service would solve this dilemma of delay where comments would be posted at the discretion of the editors, who should apply liberal policy in screening, yet censor abusive, libelous, redundant, or extremely tangential comments.
- Winker MA, Fontanarosa PB. Letters: a forum for scientific discourse. JAMA 1999;281(16):1543
- Altman DG. Poor quality medical research. What can journals do? JAMA. 2002;287:2765-7
- Mayberry JF. Letters to editor. I read with interest. Postgraduate Medical Journal 2004;80:559
- Boyton RJ, Arnold PC. An audit of the BMJ’s correspondence columns. BMJ. 1990;301(6766):1419-20
- Casewell A. Letters to the editor 1991. An audit of the MJA’s correspondence column. Med J Aus 1992;157(1):63-4
- Cheng TO. Old letters, new rules. Lancet 2003;361:706
- Mullan Z. lancet correspondence: old letters, new rules. Lancet 2003;361:12
|1||scrutinize flaw in methods||3||3.95|
|2||raise new issue(s)/ suggest new references||12||15.79|
|3||Add new information||25||32.89|
|4||riticize statistical analysis or require appropriate one||1||1.32|
|5||provide alternative interpretation/ clarify contested points||4||5.26|
|6||discussion did not explain results or criticize conclusion reached||13||17.11|
|7||pose important question(s)||2||2.63|
|8||give his own or country experience||16||21.05|
|* Total exceeds the number of critical letters (52) due to multiple responses in some letters.|
Mustafa Afifi. Department of Non-Communicable Diseases, Directorate General of Health Affairs, Ministry of Health (H.Q.), P.O. Box 393, P.C. 113 Muscat, Oman. (Email: firstname.lastname@example.org)