Analysis articles are narrative evidence based essays, with or without data, which aim to foster interest and provoke comment and debate on important health issues. They should be aimed at a general medical audience which includes doctors of all disciplines and other health professionals working in and outside the UK. You should not assume that readers will know about organisations or practices that are specific to one country or a single discipline. Analysis articles should include: 

• 1700-2000 words set out under informative subheadings. Please include a 100-150 word introduction spelling out what the paper is about and emphasising its importance
• explicit evidence to support key statements and brief explanation of the strength of the evidence (published trials, systematic reviews, observational studies, expert opinion etc)
• no more than 20 references in Vancouver style, presenting the evidence on which the key statements in the paper are made
• up to three tables, boxes, or illustrations (clinical photographs, imaging, line drawings, figures - we welcome colour)  to enhance the text and add to or substantiate key points made in the body of the article
• a summary box with up to four short single sentences, in the form of bullet points, highlighting the article's main points
• a box of linked information such as website urls for those who want to pursue the subject in more depth: this is optional
• web extras: we may be able to publish on bmj.com some additional boxes, figures, and references (in a separate reference list numbered w1,w2,w3 etc and marked as such in the main text of the article)
• a statement of sources and selection criteria: as well as the standard statements of funding, competing interests, and contributorship, please provide at the end of the paper a 100-150 word paragraph (excluded from word count) explaining the paper's provenance. This should include the relevant experience/expertise of the authors and the sources of information used to prepare the paper. It should also give details of each author's role in producing the article and name one as guarantor.
  
  All our research shows that readers are more likely to read an article that includes images. If you cannot supply one or more images (or ideas for images) we reserve, therefore, the right to source one or more ourselves. These will usually be photographs, but we will occasionally commission illustrations. We try to let authors know when we commission images, but this is not always possible because of time constraints when we are putting pages together.