This section will focus on suspect manuscripts that have been submitted to journals but not yet published. The section will review the issues of how a manuscript might come to be considered suspect, the steps that should be taken when misconduct is alleged, people or institutions the journal might notify about a suspect manuscript, and responsibility for investigating allegations of misconduct. Sample correspondence related to this topic is available on the CSE website.1
A number of cases involving allegations of misconduct and manuscripts have been reported. Cases also exist in which an allegation of misconduct was made even before the manuscript had been submitted to a journal. For example, even showing a draft of a manuscript that contains falsified data to collaborators may serve as the basis of a misconduct allegation. In addition to the advice rendered by the Office of Research Integrity (ORI) and the National Science Foundation (NSF), the Committee on Publication Ethics (COPE)2 provides advice to journal editors regarding the handling of suspect manuscripts.
Suspect manuscripts might be identified through any of the following means:
A number of parties can identify a manuscript whose content or authorship may reflect misconduct (herein termed a suspect manuscript). These parties include:
Addressing allegations of misconduct is a sensitive matter, and should be done with great care, as it can affect an author’s career. In correspondence concerning allegations of misconduct, it is important that language be non-accusatory, yet clear. It may be helpful to set deadlines for responses. Sample correspondence is available on the CSE website.1 In general, when a manuscript under consideration by a journal is identified as suspect, the following steps are taken:
If he or she suspects an article contains material that may result in a finding of misconduct, the editor may choose to notify some or all of the following parties:
Some editors choose to notify the corresponding author of a problem with a manuscript as recommended by some of the COPE flow charts,2 while others contact all the authors when misconduct has been alleged in a manuscript, as recommended by the ICMJE. Both approaches have advantages. The COPE approach identifies a potential problem without initiating the steps required in a misconduct investigation and it minimizes potential unnecessary harm to an author. The corresponding author often can identify which author is responsible for the suspect portion of the manuscript without unnecessarily involving the other authors and provide the corresponding author an opportunity to obtain records before a co-author can destroy them; since the corresponding author knows who is responsible for each component of the manuscript, communication with the responsible author should be sufficient without involving all authors. Limitations of this approach are that the corresponding author is not always the senior author and that it does not allow the other authors to know that their manuscript has been questioned and may be investigated.
Communicating with all authors has the benefit of allowing all the authors to police themselves and is the most transparent method. Editors following the ICMJE recommendation and contacting all of the authors when misconduct is alleged will likely receive a prompt response, but this approach potentially increases the risk that word of the inquiry might not be kept confidential or that one inquiry will result in different responses from multiple authors and institutions (for example, one institution might require the reporting of potential allegations of misconduct, while another institution might wait until a formal allegation is made). Also, authors who are not responsible for the suspect portion of the manuscript are more likely to invoke protective processes to prevent the opening of investigations at their institutions upon receiving a letter from a journal editor. Authors may also attempt to destroy or discard evidence and thus inhibit the ability of institutional authorities to resolve the issue. Authors accused of misconduct may contact the journal office about a suspect manuscript as required by their institution, funding agency, or ORI. Often, an accused author is required by his or her institution to send notice to a journal to withdraw a manuscript after an allegation is made. The notice to the journal typically does not indicate that the manuscript is the subject of a misconduct investigation. As a condition of settlement, or as a sanction imposed after a finding of misconduct, the ORI requires an accused author to send notification to a journal requesting appropriate corrective action with respect to a suspect manuscript. Other, unaccused authors may provide such notice if an accused author hesitates to do so.
If the author’s response is not satisfactory, many editors notify the employing institution, because the institution typically will have access to the source material, the means to conduct an investigation, the ability to compel an author’s participation in the investigation, and the ability to impose sanctions. However, notifying an author’s institution should not be a reflex reaction for editors. Editors should consider the impact such notification may have on the career of the accused. Relatively few editors opt to notify the relevant federal agency (where suspect manuscripts are supported by federal funding), because the jurisdiction of the agencies is often unclear when a manuscript is submitted and because the agencies will likely refer the matter back to the employing institution for investigation. Also, notification of a federal agency places the journal in the role of accuser and involves the journal in the misconduct investigation, regardless of whether it wants to participate.
An editor may instruct the author of a suspect manuscript to withdraw the submission. While this action may appear to end the problem for the current journal, the editor is effectively passing the issue on to the next (unsuspecting) journal and editor.
Few editors undertake investigations into misconduct allegations themselves. Journals often lack access to the necessary materials or resources to conduct an investigation, and most have not adopted a definition of misconduct or established policies and procedures for conducting such investigations. Further, few editors have experience or expertise in conducting such investigations or in the nuances of the various definitions of misconduct being used by the scientific community. Because a finding of scientific misconduct typically has profound professional implications for a researcher, a journal conducting an investigation should anticipate various challenges, including legal challenges. Simply contacting the submitting author’s institution, employer, or funding body may trigger an investigation by these parties. If an investigation is pursued, journal editors may or may not be informed of the results of such investigations.
If an editor notifies an institution or federal agency of an alleged instance of misconduct, the editor should be mindful that providing any material that should be confidential may influence the institution’s or agency’s expectations and make it more difficult for the editor to withhold other confidential material as an investigation progresses. The initial letter to the institution or agency should provide a summary of the allegation. If the editor chooses to provide the actual message of accusation, however, the person who made the allegation needs to provide permission for the editor to do so.
Sometimes an anonymous source accuses an author or group of misconduct, plagiarism, or other breech of publication ethics. Making such accusations without backing them up with sufficient scientific argument or thought is, in itself, unethical in that addressing these accusations consumes considerable time on the part of editors and editorial staff, may result in unnecessary or erroneous retractions or expressions of concern, and is potentially slanderous. The NIH Committee on Scientific Conduct and Ethics states “In order to bring a formal complaint, allegations of research misconduct must be made in writing and contain sufficient details to make clear the nature of the activity and a description of the facts, events, and circumstances that led to the allegation. The signed allegation document is sent to the Agency Intramural Research Integrity Official (AIRIO). The identity of the complainant may remain confidential unless the allegations lead to an inquiry.”3
If editors/staff are notified of potential allegations of misconduct, whether or not the accusations are signed, editors/staff are obligated to look into the allegation. In the case of allegations that are not signed and are not backed by strong facts, editors have few facts to pursue the truth and cannot prove the allegations are true. If accusations are not signed, but backed by strong facts, such as proof that articles have been plagiarized as reflected by comparisons of articles with others in a public database, it is possible for editors and staff to corroborate misbehavior. In situations in which articles have been proven by such means to be plagiarized, editors may decide to publish notices of duplicate or plagiarized articles, noting that knowledge from public databases has revealed the breaches of publication ethics. Editors may be hampered in notifying authors, and imposing sanctions may not be appropriate, if addresses are out of date, authors are deceased, or authors can no longer be located.
(Authorship: Debra Parrish took the lead in writing this section of the white paper on behalf of the CSE Editorial Policy Committee. Debra Parrish revised this section for the 2009 Update. Elizabeth Blalock and Patricia Baskin revised this section for the 2012 Update. Members of the Editorial Policy Committee and the CSE Board of Directors reviewed and commented on it. This section was formally approved by the CSE Board of Directors on March 30, 2012.)