Professional Documents
Culture Documents
2004 Urm
2004 Urm
(ICMJE), which meets annually. The ICMJE gradually has ity and clarity of reporting in manuscripts submitted to
broadened its concerns to include ethical principles related any journal, as well as the ease of editing. At the same time,
to publication in biomedical journals. every journal has editorial requirements uniquely suited to
The ICJME has produced multiple editions of the its purposes. Authors therefore need to become familiar
Uniform Requirements for Manuscripts Submitted to Bio- with the specific instructions to authors published by the
medical Journals. Over the years, issues have arisen that go journal they have chosen for their manuscript—for exam-
This archived document is no longer current. The current document is available at www.icmje.org.
beyond manuscript preparation, resulting in the develop- ple, the topics suitable for that journal, and the types of
ment of a number of Separate Statements on editorial pol- papers that may be submitted (for example, original arti-
icy. The entire Uniform Requirements document was re- cles, reviews, or case reports)—and should follow those
vised in 1997; sections were updated in May 1999 and instructions. The Mulford Library at the Medical College
May 2000. In May 2001, the ICMJE revised the sections of Ohio maintains a useful compendium of instructions to
related to potential conflict of interest. In 2003, the com- authors at www.mco.edu/lib/instr/libinsta.html.
mittee revised and reorganized the entire document and
incorporated the Separate Statements into the text. The
committee prepared this revision in 2004.
II. ETHICAL CONSIDERATIONS IN THE CONDUCT AND
and editorial decision making. Conflict of interest exists not enter into agreements that interfere with their access to
when an author (or the author’s institution), reviewer, or the data and their ability to analyze it independently, to
editor has financial or personal relationships that inappro- prepare manuscripts, and to publish them. Authors should
priately influence (bias) his or her actions (such relation- describe the role of the study sponsor(s), if any, in study
ships are also known as dual commitments, competing in- design; in the collection, analysis, and interpretation of data;
terests, or competing loyalties). These relationships vary in the writing of the report; and in the decision to submit the
This archived document is no longer current. The current document is available at www.icmje.org.
from those with negligible potential to those with great report for publication. If the supporting source had no such
potential to influence judgment, and not all relationships involvement, the authors should so state. Biases potentially
represent true conflict of interest. The potential for conflict introduced when sponsors are directly involved in research
of interest can exist whether or not an individual believes are analogous to methodological biases of other sorts. Some
that the relationship affects his or her scientific judgment. journals, therefore, choose to include information about
Financial relationships (such as employment, consultan- the sponsor’s involvement in the methods section.
cies, stock ownership, honoraria, paid expert testimony) Editors may request that authors of a study funded by
are the most easily identifiable conflicts of interest and the an agency with a proprietary or financial interest in the
most likely to undermine the credibility of the journal, the outcome sign a statement such as, “I had full access to all
authors, and of science itself. However, conflicts can occur of the data in this study and I take complete responsibility
for other reasons, such as personal relationships, academic for the integrity of the data and the accuracy of the data
competition, and intellectual passion. analysis.” Editors should be encouraged to review copies of
All participants in the peer review and publication the protocol and/or contracts associated with project-spe-
process must disclose all relationships that could be viewed cific studies before accepting such studies for publication.
as presenting a potential conflict of interest. Disclosure of Editors may choose not to consider an article if a sponsor
these relationships is also important in connection with has asserted control over the authors’ right to publish.
editorials and review articles, because it is can be more
difficult to detect bias in these types of publications than in
reports of original research. Editors may use information II.D.3. Potential Conflicts of Interest Related to Commitments
disclosed in conflict of interest and financial interest state- of Editors, Journal Staff, or Reviewers
ments as a basis for editorial decisions. Editors should pub- Editors should avoid selecting external peer reviewers
lish this information if they believe it is important in judg- with obvious potential conflicts of interest, for example,
ing the manuscript. those who work in the same department or institution as
any of the authors. Authors often provide editors with the
II.D.1. Potential Conflicts of Interest Related to Individual names of persons they feel should not be asked to review a
Authors’ Commitments manuscript because of potential conflicts of interest, usu-
When authors submit a manuscript, whether an article ally professional. When possible, authors should be asked
or a letter, they are responsible for disclosing all financial to explain or justify their concerns; that information is impor-
and personal relationships that might bias their work. To tant to editors in deciding whether to honor such requests.
prevent ambiguity, authors must state explicitly whether Reviewers must disclose to editors any conflicts of in-
potential conflicts do or do not exist. Authors should do so terest that could bias their opinions of the manuscript, and
in the manuscript on a conflict of interest notification page they should disqualify themselves from reviewing specific
that follows the title page, providing additional detail, if manuscripts if they believe it to be appropriate. As in the
necessary, in a cover letter that accompanies the manu- case of authors, silence on the part of reviewers concerning
script. (See Section IV.A.3. Conflict of Interest Notification Page) potential conflicts may mean either that such conflicts exist
Investigators must disclose potential conflicts to study that they have failed to disclose, or that conflicts do not
participants and should state in the manuscript whether exist. Reviewers must therefore also be asked to state ex-
they have done so. plicitly whether conflicts do or do not exist. Reviewers must
Editors also need to decide when to publish informa- not use knowledge of the work, before its publication, to
tion disclosed by authors about potential conflicts. If doubt further their own interests.
exists, it is best to err on the side of publication. Editors who make final decisions about manuscripts
must have no personal, professional, or financial involve-
II.D.2. Potential Conflicts of Interest Related to Project Support ment in any of the issues they might judge. Other mem-
Increasingly, individual studies receive funding from bers of the editorial staff, if they participate in editorial
commercial firms, private foundations, and government. decisions, must provide editors with a current description
The conditions of this funding have the potential to bias of their financial interests (as they might relate to editorial
and otherwise discredit the research. judgments) and disqualify themselves from any decisions
Scientists have an ethical obligation to submit credit- where they have a conflict of interest. Editorial staff must
able research results for publication. Moreover, as the per- not use the information gained through working with
sons directly responsible for their work, researchers should manuscripts for private gain. Editors should publish regu-
4
Uniform Requirements for Manuscripts Submitted to Biomedical Journals: Writing and Editing for Biomedical Publication ICMJE
lar disclosure statements about potential conflicts of inter- Reviewer comments should not be published or oth-
ests related to the commitments of journal staff. erwise made public without permission of the reviewer,
author, and editor.
II.E. Privacy and Confidentiality
Opinions differ on whether reviewers should remain
II.E.1. Patients and Study Participants
anonymous. Authors should consult the information for
Patients have a right to privacy that should not be authors of the journal they have chosen to learn whether
This archived document is no longer current. The current document is available at www.icmje.org.
infringed without informed consent. Identifying informa- the reviews are anonymous. When comments are not
tion, including patients’ names, initials, or hospital num- signed the reviewers’ identity must not be revealed to the
bers, should not be published in written descriptions, pho- author or anyone else without the reviewer’s permission.
tographs, and pedigrees unless the information is essential Some journals publish reviewers’ comments with the
for scientific purposes and the patient (or parent or guard- manuscript. No such procedure should be adopted without
ian) gives written informed consent for publication. In- the consent of the authors and reviewers. However, review-
formed consent for this purpose requires that a patient who ers’ comments should be sent to other reviewers of the same
is identifiable be shown the manuscript to be published. manuscript, which helps reviewers learn from the review pro-
Identifying details should be omitted if they are not cess, and reviewers may be notified of the editor’s decision.
essential. Complete anonymity is difficult to achieve, how-
ever, and informed consent should be obtained if there is II.F. Protection of Human Subjects and Animals in
any doubt. For example, masking the eye region in photo- Research
graphs of patients is inadequate protection of anonymity. If When reporting experiments on human subjects, au-
identifying characteristics are altered to protect anonymity, thors should indicate whether the procedures followed
such as in genetic pedigrees, authors should provide assur- were in accordance with the ethical standards of the re-
ance that alterations do not distort scientific meaning and sponsible committee on human experimentation (institu-
editors should so note. tional and national) and with the Helsinki Declaration of
The requirement for informed consent should be in- 1975, as revised in 2000 (5). If doubt exists whether the
cluded in the journal’s instructions for authors. When in- research was conducted in accordance with the Helsinki
formed consent has been obtained it should be indicated in Declaration, the authors must explain the rationale for
the published article. their approach, and demonstrate that the institutional re-
view body explicitly approved the doubtful aspects of the
study. When reporting experiments on animals, authors
II.E.2. Authors and Reviewers
should be asked to indicate whether the institutional and
Manuscripts must be reviewed with due respect for national guide for the care and use of laboratory animals
authors’ confidentiality. In submitting their manuscripts was followed.
for review, authors entrust editors with the results of their
scientific work and creative effort, on which their reputa-
III. PUBLISHING AND EDITORIAL ISSUES RELATED TO
tion and career may depend. Authors’ rights may be vio-
lated by disclosure of the confidential details of the review PUBLICATION IN BIOMEDICAL JOURNALS
of their manuscript. Reviewers also have rights to confiden- III.A. Obligation to Publish Negative Studies
tiality, which must be respected by the editor. Confidenti- Editors should consider seriously for publication any
ality may have to be breached if dishonesty or fraud is carefully done study of an important question, relevant to
alleged but otherwise must be honored. their readers, whether the results are negative (that is, con-
Editors must not disclose information about manu- vincingly allow the null hypothesis to be accepted) or pos-
scripts (including their receipt, content, status in the re- itive (that is, allow the null hypothesis to be rejected).
viewing process, criticism by reviewers, or ultimate fate) to Failure to submit or publish negative studies, in particular,
anyone other than the authors and reviewers. This includes contributes to publication bias. Many studies that purport
requests to use the materials for legal proceedings. to be negative are, in fact, inconclusive; publication of in-
Editors must make clear to their reviewers that manu- conclusive studies is problematic, since they add little to
scripts sent for review are privileged communications and biomedical knowledge and consume journal resources. The
are the private property of the authors. Therefore, review- Cochrane Library may be interested in publishing incon-
ers and members of the editorial staff must respect the clusive trials (www.cochrane.org).
authors’ rights by not publicly discussing the authors’ work III.B. Corrections, Retractions and “Expressions of
or appropriating their ideas before the manuscript is pub- Concern”
lished. Reviewers must not be allowed to make copies of Editors must assume initially that authors are report-
the manuscript for their files and must be prohibited from ing work based on honest observations. Nevertheless, two
sharing it with others, except with the permission of the types of difficulty may arise.
editor. Reviewers should return or destroy copies of manu- First, errors may be noted in published articles that
scripts after submitting reviews. Editors should not keep require the publication of a correction or erratum of part of
copies of rejected manuscripts. the work. The corrections should appear on a numbered
5
ICMJE Uniform Requirements for Manuscripts Submitted to Biomedical Journals: Writing and Editing for Biomedical Publication
page, be listed in the contents page, include the complete under serial rights (that is, use in publications other than
original citation, and link to the original article and vice journals, including electronic publications, is permitted).
versa if online. It is conceivable that an error could be so
III.D. Overlapping Publications
serious as to vitiate the entire body of the work, but this is
III.D.1. Duplicate Submission
unlikely and should be handled by editors and authors on
Most biomedical journals will not consider manu-
an individual basis. Such an error should not be confused
This archived document is no longer current. The current document is available at www.icmje.org.
enced in the new paper. Copies of such material should be cate that it is a secondary publication (complete repub-
included with the submitted paper to help the editor de- lication, abridged republication, complete translation,
cide how to handle the matter. or abridged translation) of a primary publication. Of note,
If redundant or duplicate publication is attempted or the National Library of Medicine does not consider trans-
occurs without such notification, authors should expect ed- lations to be “republications,” and does not cite or index
itorial action to be taken. At the least, prompt rejection of translations when the original article was published in a
This archived document is no longer current. The current document is available at www.icmje.org.
the submitted manuscript should be expected. If the editor journal that is indexed in MEDLINE.
was not aware of the violations and the article has already
been published, then a notice of redundant or duplicate III.D.4. Competing Manuscripts Based on the Same Study
publication will probably be published with or without the Publication of manuscripts to air co-investigators dis-
author’s explanation or approval. putes may waste journal space and confuse readers. On the
Preliminary reporting to public media, governmental other hand, if editors knowingly publish a manuscript
agencies, or manufacturers, of scientific information de- written by only some of a collaborating team, they could
scribed in a paper or a letter to the editor that has been be denying the rest of the team their legitimate co author-
accepted but not yet published violates the policies of ship rights; they could also be denying the journal’s readers
many journals. Such reporting may be warranted when the access to legitimate differences of opinion about the inter-
paper or letter describes major therapeutic advances or pretation of a study.
public health hazards such as serious adverse effects of Two kinds of competing submissions are considered:
drugs, vaccines, other biological products, or medicinal de- submissions by coworkers who disagree on the analysis and
vices, or reportable diseases. This reporting should not interpretation of their study, and submissions by coworkers
jeopardize publication, but should be discussed with and who disagree on what the facts are and which data should
agreed upon by the editor in advance. be reported.
Setting aside the unresolved question of ownership of
the data, the following general observations may help edi-
III.D.3. Acceptable Secondary Publication tors and others dealing with these problems.
Certain types of articles, such as guidelines produced
by governmental agencies and professional organizations,
III.D.4.a. Differences in Analysis or Interpretation
may need to reach the widest possible audience. In such
If the dispute centers on the analysis or interpretation
instances, editors sometimes choose deliberately to publish
of data, the authors should submit a manuscript that
material that is also being published in other journals, with
clearly presents both versions. The difference of opinion
the agreement of the authors and the editors of those other
should be explained in a cover letter. The normal process
journals. Secondary publication for various other reasons,
of peer and editorial review of the manuscript may help the
in the same or another language, especially in other coun-
authors to resolve their disagreement regarding analysis or
tries, is justifiable, and can be beneficial, provided all of the
interpretation.
following conditions are met.
If the dispute cannot be resolved and the study merits
1. The authors have received approval from the editors
publication, both versions should be published. Options
of both journals; the editor concerned with secondary pub-
include publishing two papers on the same study, or a
lication must have a photocopy, reprint, or manuscript of
single paper with two analyses or interpretations. In such
the primary version.
cases it would be appropriate for the editor to publish a
2. The priority of the primary publication is respected
statement outlining the disagreement and the journal’s in-
by a publication interval of at least one week (unless spe-
volvement in attempts to resolve it.
cifically negotiated otherwise by both editors).
3. The paper for secondary publication is intended for
a different group of readers; an abbreviated version could III.D.4.b. Differences in Reported Methods or Results
be sufficient. If the dispute centers on differing opinions of what
4. The secondary version faithfully reflects the data was actually done or observed during the study, the journal
and interpretations of the primary version. editor should refuse publication until the disagreement is
5. The footnote on the title page of the secondary resolved. Peer review cannot be expected to resolve such
version informs readers, peers, and documenting agencies problems. If there are allegations of dishonesty or fraud,
that the paper has been published in whole or in part and editors should inform the appropriate authorities; authors
states the primary reference. A suitable footnote might should be notified of an editor’s intention to report a sus-
read: “This article is based on a study first reported in the picion of research misconduct.
[title of journal, with full reference].”
Permission for such secondary publication should be III.D.5. Competing Manuscripts Based on the Same Database
free of charge. Editors sometimes receive manuscripts from separate
6. The title of the secondary publication should indi- research groups that have analyzed the same data set, e.g.,
7
ICMJE Uniform Requirements for Manuscripts Submitted to Biomedical Journals: Writing and Editing for Biomedical Publication
from a public database. The manuscripts may differ in their improved cooperation between academic and corporate en-
analytic methods, conclusions, or both. Each manuscript tities. Because funding sources can bias the content of sup-
should be considered separately. Where interpretations of plements through the choice of topics and viewpoints,
the same data are very similar, it is reasonable but not journals should consider adopting the following principles.
necessary for editors to give preference to the manuscript These same principles apply to theme issues or special se-
that was received earlier. However, editorial consideration ries that have external funding and/or guest editors.
This archived document is no longer current. The current document is available at www.icmje.org.
of multiple submissions may be justified in this circum- 1. The journal editor must take full responsibility for
stance, and there may even be a good reason for publishing the policies, practices, and content of supplements, includ-
more than one manuscript because different analytical ap- ing complete control of the decision to publish all portions
proaches may be complementary and equally valid. of the supplement. Editing by the funding organization
III.E. Correspondence
should not be permitted.
2. The journal editor must retain the authority to send
Biomedical journals should provide its readership with
supplement manuscripts for external peer review and to
a mechanism for submitting comments, questions, or crit-
reject manuscripts submitted for the supplement. These
icisms about published articles, as well as brief reports and
conditions should be made known to authors and external
commentary unrelated to previously published articles.
supplement editors before beginning editorial work on the
This will likely, but not necessarily, take the form of a
supplement.
correspondence section or column. The authors of articles
3. The journal editor must approve the appointment
discussed in correspondence should be given an opportu-
of any external editor of the supplement and take respon-
nity to respond, preferably in the same issue in which the
original correspondence appears. Authors of correspon- sibility for the work of the external editor.
dence should be asked to declare any competing or con- 4. The sources of funding for the research, publica-
flicting interests. tion, and the products the funding source make that are
Published correspondence may be edited for length, considered in the supplement should be clearly stated and
grammatical correctness, and journal style. Alternatively, prominently located in the supplement, preferably on each
editors may choose to publish correspondence unedited for page. Whenever possible, funding should come from more
length or style, as for example in rapid response sections on than one sponsor.
the Internet; the journal should declare its editorial practice 5. Advertising in supplements should follow the same
in this regard. Authors should approve editorial changes policies as those of the rest of the journal.
that alter the substance or tone of a letter or response. 6. Journal editors must enable readers to distinguish
Although editors have the prerogative to sift out cor- readily between ordinary editorial pages and supplement
respondence material that is irrelevant, uninteresting, or pages.
lacking in cogency, they have a responsibility to allow a 7. Journal editors and supplement editors must not
range of opinion to be expressed. The correspondence col- accept personal favors or personal remuneration from
umn should not be used merely to promote the journal’s, sponsors of supplements.
or the editors’, point of view. In all instances, editors must 8. Secondary publication in supplements (republica-
make an effort to screen out discourteous, inaccurate, or tion of papers previously published elsewhere) should be
libelous statements, and should not allow ad hominem ar- clearly identified by the citation of the original paper. Sup-
guments intended to discredit opinions or findings. plements should avoid redundant or duplicate publication.
In the interests of fairness and to keep correspondence Supplements should not republish research results, but the
within manageable proportions, journals may want to set republication of guidelines or other material in the public
time limits for responding to articles and correspondence, interest might be appropriate.
and for debate on a given topic. Journals should also decide 9. The principles of authorship and potential conflict
whether they would notify authors when correspondence of interest disclosure articulated elsewhere in this docu-
bearing on their published work is going to appear in stan- ment should apply to supplements.
dard or rapid response sections. Journals should also set III.G. Electronic Publishing
policy with regard to the archiving of unedited correspon-
Most biomedical journals are now published in elec-
dence that appears on line. These policies should be pub-
tronic as well as print versions, and some are published in
lished both in print and electronic versions of the journal.
electronic form only. Electronic publishing (which in-
III.F. Supplements, Theme Issues, and Special Series cludes the Internet) is publishing. In the interests of clarity
Supplements are collections of papers that deal with and consistency, the medical and health information pub-
related issues or topics, are published as a separate issue of lished on the Internet should follow the recommendations
the journal or as part of a regular issue, and are usually in this document whenever possible.
funded by sources other than the journal’s publisher. Sup- The nature of electronic publication requires some
plements can serve useful purposes: education, exchange of special considerations, both within and beyond this docu-
research information, ease of access to focused content, and ment. At a minimum, websites should indicate the fol-
8
Uniform Requirements for Manuscripts Submitted to Biomedical Journals: Writing and Editing for Biomedical Publication ICMJE
lowing: names, appropriate credentials, affiliations, and where required by law. Finally, editors should consider all
relevant conflicts of interest of editors, authors, and con- criticisms of advertisements for publication.
tributors; documentation and attribution of references and
sources for all content; information about copyright; dis- III.I. Medical Journals and the General Media
closure of site ownership; and disclosure of sponsorship, The public’s interest in news of medical research has
advertising, and commercial funding. led the popular media to compete vigorously to get infor-
This archived document is no longer current. The current document is available at www.icmje.org.
Linking from one health or medical Internet site to mation about research as soon as possible. Researchers and
another may be perceived as an implicit recommendation institutions sometimes encourage the reporting of research
of the quality of the second site. Journals thus should ex- in the non-medical media before full publication in a sci-
ercise caution in linking to other sites; when users are link- entific journal by holding a press conference or giving
ing to another site, it may be helpful to provide an explicit interviews.
message to that they are leaving the journal’s site. If links to The public is entitled to important medical informa-
other sites are posted as a result of financial considerations, tion without unreasonable delay, and editors have a re-
such should be clearly indicated. All dates of content post- sponsibility to play their part in this process. Biomedical
ing and updating should be indicated. In electronic layout journals are published primarily for their readers, but the
as in print, advertising and promotional messages should general public has a legitimate interest in their content; an
not be juxtaposed with editorial content, and commercial appropriate balance should therefore guide journals’ inter-
content should be clearly identifiable as such. action with the media between these complementary inter-
Electronic publication is an area that is in flux. Editors ests. Doctors in practice need to have reports available in
should develop, make available to authors, and implement full detail before they can advise their patients about the
policies on issues unique to electronic publishing. These reports’ conclusions. Moreover, media reports of scientific
issues include archiving, error correction, version control, research before the work has been peer reviewed and fully
and choice of the electronic or print version of the journal published may lead to the dissemination of inaccurate or
as the journal of record, publication of ancillary material, premature conclusions.
and electronic publication. An embargo system has been established in some
countries to prevent publication of stories in the general
III.H. Advertising media before the original paper on which they are based
Most medical journals carry advertising, which gener- appears in the journal. The embargo creates a “level play-
ates income for their publishers, but advertising must not ing field,” which most reporters appreciate since it mini-
be allowed to influence editorial decisions. Journals should mizes the pressure on them to publish stories which they
have formal, explicit, written policies for advertising in have not had time to prepare carefully. Consistency in the
both print and electronic versions; website advertising pol- timing of public release of biomedical information is also
icy should parallel policy for the print version as much as important in minimizing economic chaos, since some arti-
possible. Editors must have full and final authority for ap- cles contain information that has great potential to influ-
proving advertisements and enforcing advertising policy. ence financial markets. On the other hand, the embargo
Where independent bodies for reviewing advertising exist system has been challenged as being self-serving of jour-
editors should make use of their judgments. nals’ interests, and impeding the rapid dissemination of
Readers should be able to distinguish readily between scientific information.
advertising and editorial material. The juxtaposition of ed- Editors may find the following recommendations use-
itorial and advertising material on the same products or ful as they seek to establish policies on these issues.
subjects should be avoided. Interleafing advertising pages
within articles discourages readers by interrupting the flow Y Editors can foster the orderly transmission of medical
of editorial content, and should be discouraged. Advertis- information from researchers, through peer-reviewed
ing should not be sold on the condition that it will appear journals, to the public. This can be accomplished by an
in the same issue as a particular article. agreement with authors that they will not publicize
Journals should not be dominated by advertising, but their work while their manuscript is under consider-
ation or awaiting publication and an agreement with
editors should be careful about publishing advertisements
the media that they will not release stories before pub-
from only one or two advertisers, as readers may perceive
lication in the journal, in return for which the journal
that these advertisers have influenced the editor. will cooperate with them in preparing accurate stories.
Journals should not carry advertisements for products
that have proved to be seriously harmful to health—for Y Editors need to keep in mind that an embargo sys-
example, tobacco. Editors should ensure that existing reg- tem works on the honor system; no formal enforce-
ulatory or industry standards for advertisements specific to ment or policing mechanism exists. The decision of any
their country are enforced, or develop their own standards. significant number of media outlets, or of biomedical
The interests of organizations or agencies should not con- journals, not to respect the embargo system would there-
trol classified and other non-display advertising, except fore lead to its rapid dissolution.
9
ICMJE Uniform Requirements for Manuscripts Submitted to Biomedical Journals: Writing and Editing for Biomedical Publication
Y Editors, authors, and the media should apply the IV.A.1.a. General Principles for Manuscript Preparation
above stated principles to material released early in elec- The text of observational and experimental articles is
tronic versions of journals. usually (but not necessarily) divided into sections with the
headings Introduction, Methods, Results, and Discussion.
III.J. Obligation to Register Clinical Trials This so-called “IMRAD” structure is not simply an arbi-
The ICMJE member journals will require, as a condi- trary publication format, but rather a direct reflection of
tion of consideration for publication, registration in a pub- the process of scientific discovery. Long articles may need
lic trials registry. Trials must register at or before the onset subheadings within some sections (especially the Results
of patient enrollment. This policy applies to any clinical and Discussion sections) to clarify their content. Other
trial starting enrollment after July 1, 2005. For trials that types of articles, such as case reports, reviews, and editori-
began enrollment prior to this date, the ICMJE member als, are likely to need other formats.
journals will require registration by September 13, 2005 Publication in electronic formats has created opportu-
before considering the trial for publication. We speak only nities for adding details or whole sections in the electronic
for ourselves, but we encourage editors of other biomedical version only, layering information, cross-linking or extract-
journals to adopt similar policies. For this purpose, the ing portions of articles, and the like. Authors need to work
ICMJE defines a clinical trial as any research project that closely with editors in developing or using such new pub-
prospectively assigns human subjects to intervention or lication formats and should submit material for potential
comparison groups to study the cause-and-effect relation- supplementary electronic formats for peer review.
ship between a medical intervention and a health outcome. Double spacing of all portions of the manuscript—
Studies designed for other purposes, such as to study phar- including the title page, abstract, text, acknowledgments,
macokinetics or major toxicity (e.g., phase I trials), would references, individual tables, and legends—and generous
be exempt. margins make it possible for editors and reviewers to edit
The ICMJE does not advocate one particular registry, the text line by line, and add comments and queries, di-
but its member journals will require authors to register rectly on the paper copy. If manuscripts are submitted
their trial in a registry that meets several criteria. The reg- electronically, the files should be double spaced, because
10
Uniform Requirements for Manuscripts Submitted to Biomedical Journals: Writing and Editing for Biomedical Publication ICMJE
the manuscript may need to be printed out for reviewing (count letters and spaces) at the foot of the title page.
and editing. Running heads are published in most journals, but are also
During the editorial process reviewers and editors fre- sometimes used within the editorial office for filing and
quently need to refer to specific portions of the manu- locating manuscripts.
script, which is difficult unless the pages are numbered. 9. Word counts. A word count for the text only (excluding
Authors should therefore number all of the pages of the abstract, acknowledgments, figure legends, and references)
This archived document is no longer current. The current document is available at www.icmje.org.
manuscript consecutively, beginning with the title page. allows editors and reviewers to assess whether the informa-
tion contained in the paper warrants the amount of space
V.A.1.b. Reporting Guidelines for Specific Study Designs devoted to it, and whether the submitted manuscript fits
Research reports frequently omit important informa- within the journal’s word limits. A separate word count for
tion. The general requirements listed in the next section the Abstract is also useful for the same reason.
relate to reporting essential elements for all study designs. 10. The number of figures and tables. It is difficult for
Authors are encouraged in addition to consult reporting editorial staff and reviewers to tell if the figures and tables
guidelines relevant to their specific research design. For that should have accompanied a manuscript were actually
reports of randomized controlled trials authors should refer included unless the numbers of figures and tables that belong
to the CONSORT statement (www.consort-statement to the manuscript are noted on the title page.
.org). This guideline provides a set of recommendations
comprising a list of items to report and a patient flow IV.A.3. Conflict of Interest Notification Page
diagram. Reporting guidelines have also been developed for To prevent the information on potential conflict of
a number of other study designs that some journals may interest for authors from being overlooked or misplaced, it
ask authors to follow. Some of these reporting guidelines is necessary for that information to be part of the manu-
can also be found at www.consort-statement.org. Authors script. It should therefore also be included on a separate
should consult the information for authors of the journal page or pages immediately following the title page. How-
they have chosen. ever, individual journals may differ in where they ask au-
thors to provide this information and some journals do not
IV.A.2. Title Page send information on conflicts of interest to reviewers. (See
The title page should carry the following information: Section II.D. Conflicts of Interest)
1. The title of the article. Concise titles are easier to read IV.A.4. Abstract and Key Words
than long, convoluted ones. Titles that are too short may, An abstract (requirements for length and structured
however, lack important information, such as study de- format vary by journal) should follow the title page. The
sign (which is particularly important in identifying ran- abstract should provide the context or background for the
domized controlled trials). Authors should include all study and should state the study’s purposes, basic proce-
information in the title that will make electronic retrieval dures (selection of study subjects or laboratory animals,
of the article both sensitive and specific. observational and analytical methods), main findings (giv-
2. Authors’ names and institutional affiliations. Some jour- ing specific effect sizes and their statistical significance, if
nals publish each author’s highest academic degree(s), possible), and principal conclusions. It should emphasize
while others do not. new and important aspects of the study or observations.
3. The name of the department(s) and institution(s) to Because abstracts are the only substantive portion of
which the work should be attributed. the article indexed in many electronic databases, and the
4. Disclaimers, if any. only portion many readers read, authors need to be careful
5. Corresponding authors. The name, mailing address, that abstracts reflect the content of the article accurately.
telephone and fax numbers, and e-mail address of the au- Unfortunately, many abstracts disagree with the text of the
thor responsible for correspondence about the manuscript article (6). The format required for structured abstracts
(the “corresponding author;” this author may or may not differs from journal to journal, and some journals use more
be the “guarantor” for the integrity of the study as a whole, than one structure; authors should make it a point prepare
if someone is identified in that role. The corresponding their abstracts in the format specified by the journal they
author should indicate clearly whether his or her e-mail have chosen.
address is to be published. Some journals request that, following the abstract, au-
6. The name and address of the author to whom requests thors provide, and identify as such, 3 to 10 key words or
for reprints should be addressed or a statement that re- short phrases that capture the main topics of the article.
prints will not be available from the authors. These will assist indexers in cross-indexing the article and
7. Source(s) of support in the form of grants, equipment, may be published with the abstract. Terms from the Med-
drugs, or all of these. ical Subject Headings (MeSH) list of Index Medicus should
8. A running head. Some journals request a short running be used; if suitable MeSH terms are not yet available for
head or foot line, usually of no more than 40 characters recently introduced terms, present terms may be used.
11
ICMJE Uniform Requirements for Manuscripts Submitted to Biomedical Journals: Writing and Editing for Biomedical Publication
IV.A.5. Introduction information about effect size. References for the design of
Provide a context or background for the study (i.e., the the study and statistical methods should be to standard
nature of the problem and its significance). State the spe- works when possible (with pages stated). Define statistical
cific purpose or research objective of, or hypothesis tested terms, abbreviations, and most symbols. Specify the com-
by, the study or observation; the research objective is often puter software used.
more sharply focused when stated as a question. Both the
This archived document is no longer current. The current document is available at www.icmje.org.
main and secondary objectives should be made clear, and IV.A.7. Results
any pre-specified subgroup analyses should be described. Present your results in logical sequence in the text,
Give only strictly pertinent references and do not include tables, and illustrations, giving the main or most important
data or conclusions from the work being reported. findings first. Do not repeat in the text all the data in the
tables or illustrations; emphasize or summarize only impor-
IV.A.6. Methods tant observations. Extra or supplementary materials and
The Methods section should include only information technical detail can be placed in an appendix where it will
that was available at the time the plan or protocol for the be accessible but will not interrupt the flow of the text;
study was written; all information obtained during the alternatively, it can be published only in the electronic
conduct of the study belongs in the Results section. version of the journal.
When data are summarized in the Results section, give
IV.A.6.a. Selection and Description of Participants numeric results not only as derivatives (for example, per-
Describe your selection of the observational or exper- centages) but also as the absolute numbers from which the
imental participants (patients or laboratory animals, in- derivatives were calculated, and specify the statistical meth-
cluding controls) clearly, including eligibility and exclusion ods used to analyze them. Restrict tables and figures to
criteria and a description of the source population. Because those needed to explain the argument of the paper and to
the relevance of such variables as age and sex to the object assess its support. Use graphs as an alternative to tables
of research is not always clear, authors should explain their with many entries; do not duplicate data in graphs and
use when they are included in a study report; for example, tables. Avoid non-technical uses of technical terms in sta-
authors should explain why only subjects of certain ages tistics, such as “random” (which implies a randomizing de-
were included or why women were excluded. The guiding vice), “normal,” “significant,” “correlations,” and “sample.”
principle should be clarity about how and why a study was Where scientifically appropriate, analyses of the data
done in a particular way. When authors use variables such by variables such as age and sex should be included.
as race or ethnicity, they should define how they measured
the variables and justify their relevance. IV.A.8. Discussion
Emphasize the new and important aspects of the study
IV.A.6.b. Technical information and the conclusions that follow from them. Do not repeat
Identify the methods, apparatus (give the manufactur- in detail data or other material given in the Introduction or
er’s name and address in parentheses), and procedures in the Results section. For experimental studies it is useful to
sufficient detail to allow other workers to reproduce the begin the discussion by summarizing briefly the main find-
results. Give references to established methods, including ings, then explore possible mechanisms or explanations for
statistical methods (see below); provide references and brief these findings, compare and contrast the results with other
descriptions for methods that have been published but are relevant studies, state the limitations of the study, and ex-
not well known; describe new or substantially modified meth- plore the implications of the findings for future research
ods, give reasons for using them, and evaluate their limita- and for clinical practice.
tions. Identify precisely all drugs and chemicals used, includ- Link the conclusions with the goals of the study but
ing generic name(s), dose(s), and route(s) of administration. avoid unqualified statements and conclusions not ade-
Authors submitting review manuscripts should include quately supported by the data. In particular, authors should
a section describing the methods used for locating, select- avoid making statements on economic benefits and costs
ing, extracting, and synthesizing data. These methods should unless their manuscript includes the appropriate economic
also be summarized in the abstract. data and analyses. Avoid claiming priority and alluding to
work that has not been completed. State new hypotheses
IV.A.6.c. Statistics when warranted, but clearly label them as such.
Describe statistical methods with enough detail to en-
able a knowledgeable reader with access to the original data IV.A.9. References
to verify the reported results. When possible, quantify find-
ings and present them with appropriate indicators of mea- IV.A.9.a. General Considerations Related to References
surement error or uncertainty (such as confidence inter- Although references to review articles can be an effi-
vals). Avoid relying solely on statistical hypothesis testing, cient way of guiding readers to a body of literature, review
such as the use of P values, which fails to convey important articles do not always reflect original work accurately.
12
Uniform Requirements for Manuscripts Submitted to Biomedical Journals: Writing and Editing for Biomedical Publication ICMJE
references can now be added to the electronic version arate sheet of paper. Number tables consecutively in the order
of published papers, and since electronic literature of their first citation in the text and supply a brief title for
searching allows readers to retrieve published literature each. Do not use internal horizontal or vertical lines. Give
efficiently. each column a short or abbreviated heading. Authors should
Avoid using abstracts as references. References to pa- place explanatory matter in footnotes, not in the heading.
pers accepted but not yet published should be designated Explain in footnotes all nonstandard abbreviations. For
as “in press” or “forthcoming”; authors should obtain writ- footnotes use the following symbols, in sequence:
ten permission to cite such papers as well as verification *,†,‡,§,储,**,††,‡‡
that they have been accepted for publication. Information
from manuscripts submitted but not accepted should be Identify statistical measures of variations, such as stan-
cited in the text as “unpublished observations” with written dard deviation and standard error of the mean.
permission from the source. Be sure that each table is cited in the text.
Avoid citing a “personal communication” unless it If you use data from another published or unpublished
provides essential information not available from a public source, obtain permission and acknowledge them fully.
source, in which case the name of the person and date of Additional tables containing backup data too extensive
communication should be cited in parentheses in the text. to publish in print may be appropriate for publication in
For scientific articles, authors should obtain written per- the electronic version of the journal, deposited with an
mission and confirmation of accuracy from the source of a archival service, or made available to readers directly by the
personal communication. authors. In that event an appropriate statement will be add-
Some journals check the accuracy of all reference cita- ed to the text. Submit such tables for consideration with
tions, but not all journals do so, and citation errors some- the paper so that they will be available to the peer reviewers.
times appear in the published version of articles. To min-
IV.A.11. Illustrations (Figures)
imize such errors, authors should therefore verify references
against the original documents. Figures should be either professionally drawn and pho-
tographed, or submitted as photographic quality digital
prints. In addition to requiring a version of the figures
suitable for printing, some journals now ask authors for
IV.A.9.b. Reference Style and Format electronic files of figures in a format (e.g., JPEG or GIF)
The Uniform Requirements style is based largely on that will produce high quality images in the web version of
an ANSI standard style adapted by the National Library of the journal; authors should review the images of such files
Medicine (NLM) for its databases. (7) For samples of ref- on a computer screen before submitting them, to be sure
erence citation formats, authors should consult http://www they meet their own quality standard.
.nlm.nih.gov/bsd/uniform_requirements.html. For x-ray films, scans, and other diagnostic images, as
References should be numbered consecutively in the well as pictures of pathology specimens or photomicro-
order in which they are first mentioned in the text. Identify graphs, send sharp, glossy, black-and-white or color pho-
references in text, tables, and legends by Arabic numerals tographic prints, usually 127 ⫻ 173 mm (5 ⫻ 7 inches).
in parentheses. References cited only in tables or figure Although some journals redraw figures, many do not. Let-
legends should be numbered in accordance with the se- ters, numbers, and symbols on Figures should therefore be
quence established by the first identification in the text of clear and even throughout, and of sufficient size that when
the particular table or figure. The titles of journals should reduced for publication each item will still be legible. Fig-
be abbreviated according to the style used in Index Medi- ures should be made as self-explanatory as possible, since
cus. Consult the List of Journals Indexed for MEDLINE, many will be used directly in slide presentations. Titles and
published annually as a separate publication by the detailed explanations belong in the legends, however, not
National Library of Medicine. The list can also be on the illustrations themselves.
obtained through the Library’s web site Photomicrographs should have internal scale markers.
(http://www.nlm.nih.gov/tsd/serials/lji.html). Symbols, arrows, or letters used in photomicrographs
Journals vary on whether they ask authors to cite elec- should contrast with the background.
tronic references within parentheses in the text or in num- If photographs of people are used, either the subjects
bered references following the text. Authors should consult must not be identifiable or their pictures must be accom-
with the journal that they plan to submit their work to. panied by written permission to use the photograph (see
13
ICMJE Uniform Requirements for Manuscripts Submitted to Biomedical Journals: Writing and Editing for Biomedical Publication
Section III.D.4.a). Whenever possible permission for pub- onto the journal website. Electronic submission saves time
lication should be obtained. as well as postage costs, and allows the manuscript to be
Figures should be numbered consecutively according handled in electronic form throughout the editorial process
to the order in which they have been first cited in the text. (for example, when it is sent out for review). When sub-
If a figure has been published, acknowledge the original mitting a manuscript electronically, authors should consult
source and submit written permission from the copyright with the instructions for authors of the journal they have
This archived document is no longer current. The current document is available at www.icmje.org.
holder to reproduce the material. Permission is required chosen for their manuscript.
irrespective of authorship or publisher except for docu- If a paper version of the manuscript is submitted, send
ments in the public domain. the required number of copies of the manuscript and fig-
For illustrations in color, ascertain whether the journal ures; they are all needed for peer review and editing, and
requires color negatives, positive transparencies, or color editorial office staff cannot be expected to make the re-
prints. Accompanying drawings marked to indicate the re- quired copies.
gion to be reproduced might be useful to the editor. Some Manuscripts must be accompanied by a cover letter,
journals publish illustrations in color only if the author which should include the following information.
pays for the extra cost.
Authors should consult the journal about require- Y A full statement to the editor about all submissions
ments for figures submitted in electronic formats. and previous reports that might be regarded as redun-
dant publication of the same or very similar work. Any
IV.A.12. Legends for Illustrations (Figures) such work should be referred to specifically, and refer-
Type or print out legends for illustrations using dou- enced in the new paper. Copies of such material should
ble spacing, starting on a separate page, with Arabic nu- be included with the submitted paper, to help the edi-
merals corresponding to the illustrations. When symbols, tor decide how to handle the matter.
arrows, numbers, or letters are used to identify parts of the
illustrations, identify and explain each one clearly in the Y A statement of financial or other relationships that
legend. Explain the internal scale and identify the method might lead to a conflict of interest, if that information
of staining in photomicrographs. is not included in the manuscript itself or in an authors’
form
IV.A.13. Units of Measurement
Measurements of length, height, weight, and volume Y A statement that the manuscript has been read and
should be reported in metric units (meter, kilogram, or approved by all the authors, that the requirements for
liter) or their decimal multiples. authorship as stated earlier in this document have been
Temperatures should be in degrees Celsius. Blood met, and that each author believes that the manuscript
pressures should be in millimeters of mercury, unless other represents honest work, if that information is not pro-
units are specifically required by the journal. vided in another form (see below); and
Journals vary in the units they use for reporting hema-
Y The name, address, and telephone number of the
tological, clinical chemistry, and other measurements. Au-
corresponding author, who is responsible for commu-
thors must consult the information for authors for the par-
nicating with the other authors about revisions and fi-
ticular journal and should report laboratory information in
nal approval of the proofs, if that information is not
both the local and International System of Units (SI). Ed-
included on the manuscript itself.
itors may request that the authors before publication add
alternative or non-SI units, since SI units are not univer-
sally used. Drug concentrations may be reported in either The letter should give any additional information that
SI or mass units, but the alternative should be provided in may be helpful to the editor, such as the type or format of
parentheses where appropriate. article in the particular journal that the manuscript repre-
sents. If the manuscript has been submitted previously to
another journal, it is helpful to include the previous edi-
IV.A.14. Abbreviations and Symbols
tor’s and reviewers’ comments with the submitted manu-
Use only standard abbreviations; the use of non-stan- script, along with the authors’ responses to those comments.
dard abbreviations can be extremely confusing to readers. Editors encourage authors to submit these previous com-
Avoid abbreviations in the title. The full term for which an munications and doing so may expedite the review process.
abbreviation stands should precede its first use in the text Many journals now provide a pre-submission checklist
unless it is a standard unit of measurement.
that assures that all the components of the submission have
IV.B Sending the Manuscript to the Journal been included. Some journals now also require that authors
An increasing number of journals now accept elec- complete checklists for reports of certain study types (e.g.,
tronic submission of manuscripts, whether on disk, as at- the CONSORT checklist for reports of randomized con-
tachments to electronic mail, or by downloading directly trolled trials). Authors should look to see if the journal uses
14
Uniform Requirements for Manuscripts Submitted to Biomedical Journals: Writing and Editing for Biomedical Publication ICMJE
such checklists, and send them with the manuscript if they VII. AUTHORS OF THE UNIFORM REQUIREMENTS FOR
are requested. MANUSCRIPTS SUBMITTED TO BIOMEDICAL JOURNALS
Copies of any permission to reproduce published ma- The ICMJE participating journals and organizations
terial, to use illustrations or report information about iden- and their representatives who approved the revised Uni-
tifiable people, or to name people for their contributions form Requirements for Manuscripts in October 2004 in-
must accompany the manuscript.
This archived document is no longer current. The current document is available at www.icmje.org.