- BACKGROUND AND RATIONALE
This policy was initially written in response to charges approved by the Medical Physics Editorial Board in 2015, which were subsequently endorsed by JACMP in 2018. This policy was drafted by the Scientific Integrity Subcommittee (SISC) of the Medical Physics Editorial Board and later updated by the Scientific Integrity Committee, whose membership was expanded in 2018 to include representation from JACMP leadership.
As listed in the AAPM Rules section 3.6.14, SIC activities pertinent to this policy are:
- Develop a written, comprehensive policy for scientific integrity standards specific to the AAPM journals, AAPM internal official reports, and their respective readership and ensure such policy is disseminated to authors and reviewers.
- Devise a process for review and adjudication of individual cases of alleged scientific integrity infractions.
In support of these activities, a policy has been prepared to define scientific integrity standards, to establish processes for review and adjudication of individual cases of alleged scientific integrity infractions, and to disseminate the requirements of this policy to Medical Physics and JACMP manuscript authors as well as editorial team members (e.g., Editors, Deputy Editors, Associate Editors, and referees) involved with the manuscript review process.
- DEFINITIONS
Scientific integrity is defined as a truthful quality of a publication or truthful actions of authors. It is expected that manuscripts submitted to Medical Physics and JACMP will exhibit scientific integrity, and the authors of such manuscripts will be truthful in their account of the manuscript contents. Examples of lapses in scientific integrity are given below.
Plagiarism is a fraudulent activity (literary theft) and can occur when the authors of a manuscript fail to inform the journal that another manuscript or publication (either by the authors or by others) contains major redundancy with the considered manuscript. Thus, plagiarism may be categorized as self plagiarism or plagiarism of others’ work.
COI is a situation in which financial or other personal considerations may compromise, or have the appearance of compromising, an author’s judgment in conducting or reporting research. In the context of scientific publishing, editorial team members can have conflicts-of-interest (COIs) with respect to their roles in the peer-review process. Often but not always, disclosing any real or apparent COIs is sufficient for management and maintenance of scientific integrity.
COI errors are situations in which an actual or apparent COI is not properly managed. Disclosure of all relevant COIs is necessary but is not always sufficient for proper management of COI. It is always better to be open.
Data fabrication is the serious act of making up data and recording or reporting these data.
Data falsification is the serious act of manipulating research materials, equipment, or processes, or changing or omitting these data or results such that the research is not accurately represented in the manuscript.
Contributor errors occur when manuscript authors and their affiliations (location of research and/or writing) are inappropriately excluded, included, or described. This results in an untruthful account of the manuscript contributors. The International Committee of Medical Journal Editors has defined the accepted requirements for authorship in biomedical journals. The individuals who conduct the work are responsible for identifying those who meet these criteria. As such, it is not the role of journals to determine who qualifies for authorship or to arbitrate conflicts of authorship contributions.
- DISCIPLINARY ACTIONS
Lapses in scientific integrity may have varying severity. The SIC will generally review potential lapses of scientific integrity when they are considered medium or high severity. An example of low-level severity is plagiarism (or self-plagiarism) such as the copying of published materials without attribution. This also includes proceedings papers, which by editorial policy are allowed to have significant overlap with Medical Physics and JACMP submissions as long as proper credit is given and cited. Another low-level example would be falsifying one’s credentials or affiliation. Medium-level examples would be at least a 2nd offense of a low-level infraction or the unacknowledged copying of data or results from other published materials such as copying from figures, tables, or whole paragraphs. High-level examples would be at least a 2nd offense of a medium-level infraction or documented data fabrication or data falsification. When scientific misconduct is reported following publication of a paper, the Editor may require publication of an erratum or retraction of the paper in addition to the author-specific actions listed below.
Lapses in scientific integrity by authors discovered during manuscript review are subject to the following consequences, based on severity score:
- low reject with opportunity to resubmit manuscript following evaluation of the authors’ response
- medium reject outright with no chance for resubmission of that manuscript
- high reject outright, place a note in the author database, and possibly contact the authors' employer
and the AAPM Ethics Committee (if any of the authors are AAPM members)
Lapses in scientific integrity by authors discovered after publication of the paper are subject to the following consequences, based on the severity score
- low note from the Editor
- medium erratum or published apology
- high retraction in whole or in part of the fully published or Early View paper, place a note in the author
database, and possibly contact the authors' employer and the AAPM Ethics Committee (if any of the
authors are AAPM members)
Lapses in scientific integrity by editorial team members during the peer-review process are less common and may be identified by the authors or Editor. The editorial team member will be informed of the suspect lapse, asked to provide an explanation for the behavior, and given a written warning. A low-level example lapse would be failure of a reviewer to disclose an apparent COI. A medium-level example would be actions to reject a competitor’s work or undue promotion of their own work such as gratuitous self citations. A high-level example would be intentional violation of reviewer or AE anonymity, or intentionally submitting a biased review in service of financial or personal interests that unfairly promotes or demotes a paper. Another high-level example would be submitting a biased review with the intent of suppressing a paper perceived to be inimical to the reviewer’s financial or academic interests. AAPM counsel will review proposed actions that involve any external interactions (e.g., contacting an employer).
Scientific integrity lapses by reviewers or AEs are subject to the following consequences:
- low removal from review of a specific manuscript, (more generally) removal from reviews of
manuscripts from a specific author group, or (following review by an Editor) retroactive declaration
of any undisclosed COI
- medium removal from all manuscripts for Medical Physics or JACMP
- high removal from all Medical Physics and JACMP manuscripts, and referral to the AAPM Ethics
Committee (if an AAPM member)
Due to their authoritative position, Editors have a special duty to avoid real or apparent COIs. This includes directing the review process for manuscripts coauthored by an Editor, an Editor's recent or current trainee, an individual who reports in/directly to an Editor, a coinvestigator on a grant supporting an Editor, a relative, or any other individual whose publication success might influence a decision that could materially benefit an Editor. Participation with authors in advisory committees, coauthoring a consensus report with an author, or past COIs that have been inactive for at least one year are generally not considered a COI.
Whenever a manuscript is received that poses a potential COI with an Editor, the Editor-in-Chief will assign it to an Editor who has no COI. Each Editor has the obligation to self-screen each assigned article for COI and to request that the Editor-in-Chief reassign the article if they identify a material COI. Accepted manuscripts can receive special recognition such as Editor’s Choice article, journal cover artwork, AAPM award nominations, special treatment such as Open Access assignment, or marketing opportunities. Editor actions to avoid real or apparent COIs include declaration of COIs or replacement with another Editor.
- PROCESS AND PROCEDURES
All lapses in scientific integrity will be recorded in a database maintained by SIC for at least 7 years. Lapses in scientific integrity by the authors or editorial team may be identified either before or after a manuscript is published.
4.1. Stages in peer-review/production workflows where lapses can be detected
There are six stages in the peer-review workflow process when lapses in scientific integrity by authors may be identified. These stages include:
- during the quality control stage performed by the Managing Editor preceding an Editor assignment,
- by the Editor/DE when considering a manuscript for review and identifying an AE,
- by the AE when considering a manuscript for review and identifying reviewers,
- by the reviewers when reviewing a manuscript for consideration for publishing in the journal,
- by the production team after a manuscript has gone through the review process, been accepted for publication, and is being typeset for proof, or
- by reviewers, readers, or colleagues of the authors following publication of a paper.
Regardless of where in the workflow a potential lapse of scientific integrity is discovered or by whom, the process for managing such incidents is the same. The potential lapse must be communicated to the Editor of record, the Editor who is managing (stages 1-4), or had managed the peer-review process (stages 5-6) of the manuscript in question who is responsible for initial investigation and disposition of the incident. For low-level potential lapses, the Editor investigates the incident to determine whether there is an actual lapse. This may involve communication with the authors such as for plagiarism or with other individuals such as for accusations of data fabrication/falsification or biased reviews. As noted above, such cases are often managed by the Editor chastising the authors and requesting a corrected or revised manuscript. For borderline medium/high level infractions, manuscript rejection without the right of appeal is often appropriate. In such cases, the Editor must include the SIC Chair on all relevant correspondence.
Alternatively, the Editor can request that the SIC evaluate the manuscript to support or invalidate the determination of a lapse in scientific integrity. Via a consultation session (or other confidential electronic means of communicating), the SIC will evaluate the manuscript and vote within 40 days of being notified of the potential infraction (with the vote determined by a simple majority of all voting members given that a quorum of SIC voting members cast votes). If SIC finds that scientific misconduct has not occurred, manuscripts in stages 1-4 of the peer review process may proceed as normal. If it is determined that there was a lapse in scientific integrity, the manuscript will be rejected using standardized language (with or without the opportunity to amend the manuscript) along with a summary of the SIC findings. The authors are allowed to appeal the finding of scientific misconduct. The Editor should request formal involvement of SIC for all cases of scientific misconduct where a disciplinary action stronger than rejection of the manuscript in question is contemplated. This includes banning the authors from submitting manuscripts to the journal, retraction of a published article (be it fully published or Early View), involving an authors’ employer or supervisor, involving the AAPM Ethics Committee, or demanding a public expression of apology for the scientific misconduct.
4.2. Plagiarism
Potential plagiarism detected in preproduction stages 1-4 is the most common form of scientific misconduct. The main tool for detecting plagiarism is CrossCheck/iThenticate software which the Managing Editor uses as a quality check to screen all new submissions to the Journal for plagiarism. The Managing Editor will flag all manuscripts with a CrossCheck score overall or for a single source higher than a specific threshold, indicating potentially substantial overlap of the manuscript with previously published materials may exist. The Editor will then review the iThenticate report in detail as the CrossCheck score alone is not diagnostic for plagiarism. For example, a manuscript may have a 20% overall CrossCheck score because of flagging numerous brief phrases of text from many different publications, erroneously including references in the comparison scope, or a very short manuscript, or because whole paragraphs were lifted from a published work. Nor is CrossCheck able to determine whether the source of duplicated material was properly credited or cited. It is common for text in Methods sections to be established and not to differ substantially for certain research topics. On the other hand, even a single figure (or figure caption) or a single paragraph that overlaps published content could be grounds for plagiarism, although it would trigger a low CrossCheck score. In addition, there are special cases where duplication of or overlap with an author’s prior publications is reasonable, e.g., dissertations, Annual Meeting abstracts, and substantive expansions of proceedings papers. Acknowledgement of these prior sources and generous use of citations is generally sufficient to avoid the charge of self-plagiarism. Duplication of large segments of text or whole figures or tables from authors’ prior publication may require permission from the publisher of the prior work. Editors typically manage low-level lapses through the EJP manuscript correction process, rejecting the paper with resubmission allowed, or addressing the lapse during the revision process. SIC should be informed of all such lapses and Editor-initiated remediations.
More severe incidents of plagiarism are less frequent. Authors have concealed a previously accepted article having substantive overlap with an accepted and published submission so that the review team would not criticize it as a duplicate publication. In such a case, retraction should be considered as a potential punitive measure. Formal involvement of SIC is indicated if evidence supports such a medium- or high-level of plagiarism in a published paper. The Editor will contact the accused author(s), the author(s) whose work may have been plagiarized, and the copyright holder of the original material if different from the author(s). These communications should include the alleged plagiarizing language, together with a copy of the original and the submitted paper, and the CrossCheck report. This information should be submitted to SIC for consideration. If all parties agree that plagiarism (whether intentional or unintentional) has occurred and SIC concurs that the severity of the infraction is sufficient, one potential remedy is a written letter of apology to be sent promptly by the offending author(s) to the Editor, the author(s), and copyright holder of the plagiarized work. Upon receipt of this apology letter, the Editor-in-Chief shall place a Notice of Plagiarism in the online version of the journal that contains the offending publication, and (if possible) in the next available printed version of the journal. This Notice of Plagiarism shall identify the offending publication, the exact text that was plagiarized, and the original publication from where the plagiarized text was extracted. Further, the offending author(s) shall agree in writing that no further dissemination of the offending publication shall occur without it being accompanied by the Notice of Plagiarism. Some authors may not issue an apology. Another option is retraction of the paper in question. This option should be reserved for cases in which research findings and data were (self)plagiarized to the extent that the scientific value of the Medical Physics or JACMP paper is called into question.
If the accused author(s) petition or deny that plagiarism has occurred, the written accusation and all supporting materials shall be referred to SIC. The accused author(s) shall be encouraged by the Editor to submit any additional information that may be relevant to defend against the accusation. Further, the accused author(s) may request in writing (and be granted) a telephone conference call (having a majority presence or SIC members) with the understanding that expenses of the accused author(s) related to the hearing shall be borne by the accused author(s). If SIC rules by majority vote in support of the accusation of plagiarism, the process described above where plagiarism is admitted shall be instituted, with or without the authors’ participation. Further, SIC shall decide whether the plagiarism is sufficiently egregious to warrant referral to the AAPM Ethics Committee if the person(s) found guilty are AAPM member(s). If SIC rules against the accusation, a letter so stating this ruling shall be provided to (as applicable) the accuser, accused author(s), author(s) of the original work, and the copyright holder if different from the author(s). It is possible that the other journal would arrive at a different conclusion, and the possibility for a competing decision should be respected.
4.3. Lapses of Scientific Integrity other than Plagiarism
Other forms of scientific misconduct that have been encountered include fabrication or falsification of data. Should this potential misconduct be discovered in pre-publication stages 1-4, the procedures outlined above for plagiarism should be followed. Upon discovery receipt of a written allegation of a potential lapse in scientific integrity in a published Medical Physics or JACMP paper, the Editor will contact the accused author(s) and any related parties. Included in the correspondence shall be the alleged infraction and any supporting materials. If all parties agree that a lapse in scientific integrity (whether intentional or unintentional) has occurred, a written letter of apology shall be sent promptly by the offending author(s) to the Editor and any related parties. Upon receipt of this apology letter, the Editors shall (depending on the nature of the infraction) request an erratum from the author(s). The erratum shall identify the offending publication and the precise nature of the scientific integrity infraction. Further, the offending author(s) shall agree in writing that no further dissemination of the offending publication shall occur without it being accompanied by the Erratum. In a situation where scientific integrity lapse calls into question the scientific value of the paper, full retraction should be considered. If the accused author(s) denies the lapse in scientific integrity, the appeal process in Section 4.1 for post-publication plagiarism incidents should be followed.
- POLICY DISSEMINATION
This policy will be posted on the Medical Physics and JACMP websites in their Instructions to Authors, and will be provided to all parties (Editorial Board members and reviewers) participating in the review processes through a link in the correspondence for invitation to participate.
|
