If you choose not to submit electronically, please submit an original double-spaced typescript and 4 copies. Please enclose a diskette with your submission containing the following information: file name, computer make, model number, operating system, word processing program, and version number.
Note that manuscripts will not be returned after the review process. Unpublished manuscripts will be destroyed.
Type—double-spaced throughout (including references)—on one side only of 8½ × 11-inch white bond paper. Number all pages. Do not justify right-hand margins. Use a printer of at least letter quality.
The length of the text of the manuscript (not including abstract, tables, figures, or references) should ideally be under 3500 words and must be no more than 4500 words.
Titles should be short, specific, and descriptive, emphasizing the main point of the article. Avoid a 2-part title, if at all possible. Do not number the title, eg, I or Part I. Do not make a declarative statement in the title. Title length, including punctuation and spaces, ideally should be under 100 characters and must not exceed 150 characters.
The title page should list full names, degrees, academic affiliations and locations of each author, and the name and address to whom reprint requests should be sent. If the paper was presented at a meeting, include the organization, place, and date of presentation. Acknowledgment of all funding support for the work should also be made on this page.
Include a structured abstract of no more than 300 words for reports of original data, reviews, and meta-analyses. (See "Preparing Structured Abstracts.") For other major manuscripts, include an unstructured abstract of no more than 200 words that summarizes the objective, main points, and conclusions of the article. Abstracts are not required for editorials, commentaries, and special features.
Conventional units of measure are preferred, with Système International (SI) units expressed secondarily (in parentheses). In tables and figures, a conversion factor to SI may be presented in the footnote or legend to economize space. Exceptions to this policy include calories, hematocrit, glycosylated hemoglobin, blood cell counts, and ejection fraction, for which conventional units alone should be expressed. The metric system is preferred for length, area, mass, and volume. (See SI conversion table.)
Use generic names of drugs, unless the specific trade name is directly relevant to the study design or discussion.
Do not use abbreviations in the title or abstract. Limit their use in the text.
Do not use footnotes or appendixes. Such materials should either be incorporated into the text or offered to interested readers on request.
Include a statistical analysis section under Methods that fully describes the application of each statistical procedure used. If a test is used that is not commonly presented in the ARCHIVES, briefly describe its purpose and how it is to be interpreted. Results should report the test statistic (eg, χ2, F or t value), degrees of freedom, and P value or confidence limits. Measures of central tendency (eg, means) should be accompanied by measures of variability (eg, SDs). In the cover letter, state the name of your statistical consultant, if appropriate.
Acknowledge all illustrations or tables adapted or reproduced from other publications and submit written permission to reproduce (in print and online) from the original publishers. (See permission form.)
Tables and figures should provide substantive data and not merely illustrate the text. While the main finding of a table or figure may be discussed in the narrative, the table or figure should not duplicate the text. Although no specific guideline can be applied to all articles, the number and length of tables and figures should be kept to a minimum.