Letters to the editor

A letter to the editor contributes a focused opinion, correction, observation or research finding to a journal’s readers. Its argument must fit limited space, meet the journal’s timing and submission rules, and remain the named author’s own view.

Purpose and audience

A journal letter addresses the readership, while the editor acts as gatekeeper and normally receives more letters than can be published. ^{1, 2} Unlike an ordinary personal letter, it must interest people beyond its recipient; unlike a newspaper letter, it usually needs good-quality evidence to support its opinion. ^{2, 3} It should be timely, concise, interesting and understandable, communicate something the writer strongly believes needs saying, and be worth reading. ^{3, 4, 5} Letters pages are often among a journal’s most widely read sections. ⁶

Four main purposes are disagreement with a published article, praise or support, a response to criticism of the writer’s own article, and rapid communication of data that may not warrant a full paper or short communication. ^{7, 8} A letter may serve several purposes, but its main focus should be clear; disagreement and replies are probably the most common. ^{7, 8} A standalone letter expressing an independent opinion must be timely, pertinent and very well written. ⁹ Responding to something recently published in the target journal is more likely to lead to publication than writing without reference to one of its articles. ¹⁰

Letters offer an entry into publication without requiring a distinguished academic position or a randomised trial; prospects can be reasonable in journals with a limited readership. ¹¹ Published letters are generally titled and indexed, making them retrievable and legitimate curriculum-vitae entries, even when not refereed; whether online rapid responses belong on a curriculum vitae is less clear. ¹¹ Some journals provide rapid online responses, including BMJ and the Journal of the American Board of Family Practice in the account published in 2018. ¹¹

A research letter can use headings such as Introduction, Methods and Results. ¹² A longer case or research article may become a letter after rejection and revision; one author’s account describes shortening a paper to 500–600 words at an editor’s suggestion. ¹³ Short publication formats still require careful writing, review by peers or a knowledgeable editor, and competition for journal space. ¹⁴

Contribution and evidence

A letter can add an omitted treatment or relevant experience, offer a different interpretation or diagnosis, challenge policy or eligibility criteria, or identify a statistical error. ^{15, 16, 17, 18, 19} Praise that adds no knowledge may waste space; one editor’s judgement was that a purely congratulatory letter from prestigious academics would have been unlikely to appear if written by an average clinician. ²⁰ Supporting facts should come from reputable sources and genuinely support the argument; disagreement with a trial based solely on unpublished personal observations is not a strong case. ^{21, 22}

The final content must accurately reflect the named author’s views, including when a medical writer or pharmaceutical company assists. ⁶ A 2007 account described companies using letters under experts’ names to encourage critical discussion of trials favourable to competitors, sometimes with writing supplied directly or indirectly by the company. ⁶

Generic commentary can fail to add enough value for limited correspondence space. ^{23, 24} In 2024, The Lancet reported growing submissions of apparently AI-generated letters comprising generic paragraphs from people without expertise in the topic; its correspondence editor would reject these for lack of useful contribution. ^{23, 24}

A 2010 letter asked for clarification of a difficult sentence about slide design; the author replied with a paraphrase explaining the intended meaning. ²⁵

A letter can seek a change in journal practice: in 2007, four AMWA Journal readers challenged an article they considered promotional and requested an advertorial policy, closing with an explicit statement of constructive intent. ²⁶ The editor thanked them, agreed that advertising should not be disguised as education and explained why the article’s placement in a professional-development section was consistent with its published purpose. ^{27, 26}

Preparation and structure

Preparation should establish the point, evidence, journal requirements and timeliness in this order. ^{28, 29}

1. Decide the main point or points and keep them simple; too many points in about 300 words cannot be communicated coherently or memorably.
2. Gather supporting data and references, using general and specialised searches for recent literature even when already familiar with the topic.
3. Check the journal’s instructions for letters, including the word limit.
4. Read examples in the target journal, because the types published vary.
5. Check topicality and timing: many journals impose a deadline for responses to earlier articles.

^{28, 29}

Before drafting a response, reread the target article, check that the criticism remains defensible and verify that the gathered evidence supports it. ²² A brief plan should set out what will be said and in what order, avoiding a rambling, unstructured letter and helping make its points clearer, easier to follow and more persuasive. ³⁰ One approach treats the letter as a tightly packaged editorial and focused literature review making a single point; another allows one or two key points. ^{31, 21, 10}

A response sequence

A response to a published paper can follow this sequence. ³¹

1. Identify and cite the paper in the first sentence, making it reference 1.
2. State why the letter is being written: agreement, disagreement, concern or another reason.
3. Give evidence, preferably from the literature rather than personal experience.
4. Conclude with a statement tying the argument together.
5. Provide the few references needed, especially for new or contrary evidence.

³¹

The sandwich

The sandwich structure offers a concise form for substantial disagreement. ³²

1. Identify the article, offer one or two complimentary sentences, then state the disagreement; the writer may also explain relevant qualifications.
2. Explain the reasons with supporting published or unpublished data, anecdotes or statistics; this is the longest section but remains concise.
3. Restate the major point and conclude succinctly.

³²

A supportive letter adding the writer’s own data can instead open directly with its main point and additional support, without fulsome preliminary praise. ³²

EPIC

An alternative, EPIC, structures either supportive or negative responses around a proposal for action and can make drafting easier, less time-consuming, concise and persuasive. ³³

1. Engage the reader with a striking fact, serious problem or unmet need.
2. Propose a specific action addressing it.
3. Illustrate how the proposal would work and why it matters, using experience or published studies.
4. Call for specific measures consistent with the proposal.

³³

Replies and corrections

A reply to criticism can acknowledge points of agreement before explaining disagreement with supporting evidence and ending with a concise restatement. ³³ The ICMJE Uniform Requirements quoted in 2007 recommended mechanisms for comments, questions and criticism of published articles and for unrelated brief reports or commentary, with authors offered a response preferably in the same issue. ⁷ This recommendation offers an opportunity to respond; the accompanying practical interpretation calls it a right of reply. ^{7, 2}

In a 2010 anaesthesia-history exchange, the correspondent acknowledged the limits of a short overview, quoted specific claims, made three historical objections and identified his own fuller book account. ^{34, 35} The authors accepted a wording correction and a regretted omission, disputed another point with reasons, and explained the space limits of their overview. ^{35, 36} In a 2011 readability exchange, a letter praised the article’s evidence-based advice while disputing one recommendation with citations; the reply agreed with almost all the criticism but supplied a practical contractual constraint. ³⁷

A 2016 correction request identified the writer’s relevant guideline expertise, quoted the requirement at issue, traced its absence through the article, allowed that the omission might be inadvertent and requested revision and a published correction. ³⁸ The authors explained how interchangeable stakeholder terms had created confusion and agreed to minor changes; the editor announced that a revised version would replace the website PDF. ^{39, 40}

Personal acquaintance with an article’s author can make praise seem suspect or criticism appear personal. ³¹ A mean-spirited or excessively clever draft should be put aside for a few days before deciding whether to send it. ³¹

Length and timing

Journal instructions should be checked before writing, including their submission deadline and limits on words, references and authors. ^{31, 28} A word limit is a ceiling, not a target. ¹⁰ A long list of agreements and disagreements can be dull and fail to communicate the main messages; the reason for writing should appear near the start and the closing sentence should summarise the opinion memorably. ¹⁰

In the JAMA rules quoted in 2018, letters about recent articles were due within four weeks, were rarely considered later, and could contain no more than 400 words, five references including the target paper, and three authors. ³¹ In the 2007 account, some journals rejected responses to articles more than three or six months old. ²⁸

AMWA Journal instructions in early 2008 limited letters commenting on journal contents to approximately 500 words or fewer and directed them to the editor. ⁴¹ Instructions from the second 2008 issue through 2009 and 2010 also required reference to contents within the preceding two issues. ^{42, 43, 44, 45, 46, 47}

Language and style

Letters are read quickly, so their structure and language should support readability. ⁴⁸ The 2007 language guidance recommends the following. ^{48, 49}

• Build paragraphs around a main sentence at the beginning and develop it logically; three or four short paragraphs can improve readability.
• Keep sentences short rather than combining several points that require rereading.
• Keep abbreviations to a minimum.
• Prefer familiar short words to pompous or pseudoscientific language: “given” rather than “administered”, and “use” rather than “utilise”.
• Cut redundant words and elaborate phrasing; “boys” can replace “male paediatric patients” and leave room to specify age.
• Avoid potentially meaningless or misunderstood buzzwords such as “proactive” and “stakeholder”.
• Distinguish results from interpretation: data show and writers suggest; “data” is plural.
• Check that a well-informed nonspecialist can understand the main points.

^{48, 49}

Submission and final checks

The final check should cover the following points. ⁵⁰

• Confirm compliance with word and reference limits.
• Read once for spelling, grammar and defined abbreviations.
• Read again for tone, readability and a logical, easy-to-follow opinion.
• Remove excessive jargon or abbreviations that reduce accessibility.
• Check citation accuracy and whether each reference supports the intended point.
• Declare conflicting interests.
• Include the writer’s full title and contact details.

⁵⁰

Affiliation should also be included. ²² Unless instructions specify otherwise, double spacing and an easily readable font can help the editor; the correct editor’s name, title and address should be used. ¹⁰ A letter should be submitted online with a cover message identifying it as a publication submission rather than private correspondence, mentioning relevant qualifications and disclosing potential conflicts. ³¹

The AMWA Journal’s Dear Edie column in 2008 requested publishable queries and comments sent directly to the columnist, preferably by email, with permission to publish and full addresses, especially city and state, for verification. ^{51, 52} Correspondence would be edited, and affiliation and other details could be published unless the columnist was told otherwise. ^{51, 52}

Editorial handling

In the BMJ letters editor’s practice summarised in 2007, letters were the readers’ forum, valued for enthusiasm, different perspectives, diversity, debate and wit. ⁵³ Letters for print were selected from unedited online rapid responses, largely for readability rather than detailed scientific criticism, and edited lightly for clarity while preserving the writer’s voice. ⁵³ The maximum was 300 words, with shorter letters encouraged; competing-interest declarations and signed patient consent were required when information from the doctor–patient relationship was described. ⁵³ In 2010, AMWA Journal reserved the right to edit letters as appropriate. ⁴⁷

A 2014 editorial view argued that signed letters should preserve the writer’s ideas and wording, with at most adjustments for grammar, brevity and clarity; its author personally preferred even grammatical errors to remain as information about the writer. ^{54, 55} The newspaper practice it described rejected discriminatory or libellous letters, corrected references, spelling, dates and facts, and shortened text without changing meaning. ⁵⁴ In the author’s own case, an appeal to the editor-in-chief restored a letter that had been extensively rewritten under his name. ⁵⁴