Statistics from Altmetric.com
Letters—an editor’s dilemma
The publication of letters within a medical journal can challenge ideas that have gone through the peer review process, correct mistakes, and initiate a dialogue between researchers and clinicians. Many important discoveries first surfaced as letters to learned journals. Indeed the idea of correspondence between researchers on a world wide basis lies at the foundation of the scientific and clinical strides that have been made across a wide range of subjects. It began to emerge as early as the 15th century when scientists across Europe exchanged ideas and challenged each other’s thinking. In our time the emergence of email has promoted the almost instant exchange of thoughts around the globe. It is against this background that many journals, including the Postgraduate Medical Journal, have fostered this exchange of ideas both in print and online. It is not surprising that so many letters should therefore start with the immortal words: “I read with interest…”.
Indeed eight of the 18 letters published over a year and half in the journal used some form of that expression. Any editor is glad to know that his or her journal is being read and that articles have prompted a response. It is customary to show the authors of the original paper the letter and seek their response. Often this will be published alongside the letter addressed to the editor. However, publication requires something more than just interest. The letter writer usually wishes to expand on the original publication or correct an error that escaped the reviewers. The writer usually takes the opportunity to present some additional data from research which he or she has published or which has no real hope of publication as a peer reviewed article. It is in these areas that there is the potential for serious problems. Where were the data published? Was it in another unreviewed letter in another journal? Would the data stand up to critical review? However, once published such letters can be cited as publications, quoted in other articles, and used in support of grant and job applications.
A recent withdrawal of a letter prompted the present review of letters published in the Postgraduate Medical Journal over an 18 month period starting with the January 2003 issue. Authors quoted themselves in 44% of letters, rising to 75% in the case of single authored letters, with most being another letter (table 1). In the 23 of 27 cases (85%) it was possible to trace the publication history of the author of a letter in the journal through PubMed. For each author all publications were sought from the database and then the abstract reviewed. The classification used in PubMed allows the ready identification of letters. This technique was limited by surnames and initials being too common to allow identification with any certainty in four cases. It also misses authors who publish under different names, such as Mayberry JF or Mayberry J. It is therefore likely that the data in table 2 underestimate the true size of the practice. Seven of the 23 authors (30%) reviewed had published five or more letters, and three (13%) more than 10 letters.
Is the publication of letters to be criticised? If they communicate new ideas or provide data to support some new hypothesis they can have a valuable role, provided that they have been subject to the same review process that more comprehensive papers must go through. This process should parallel the journal’s approach to both original research and review. If this is not the case then letters published in journals of repute have the potential to acquire a respect which they may not warrant. Against this background the editorial board of the Postgraduate Medical Journal has decided to stop publication of letters in the hard copy version of the journal, but to encourage their publication online in the web based version. However, such online publication will not be citeable. It should therefore stimulate debate but not add to an individual’s list of publications. It should also help ensure that research that has not been peer reviewed is not published. It will not, however, necessarily prevent fraud.
Discussion and exchange of ideas are fundamental to scientific research and progress. Letters can form an important aspect of the development of such ideas. They should not be a method for avoiding peer review or the creation of a large list of publications. The Postgraduate Medical Journal is committed to promoting the best practice in clinical research and education.