Scientific Misconduct—Insights From the Work of an Ethics Committee

Abstract

Scientific misconduct does not only relate to falsifying scientific data or plagiarism but may also include improper handling of authorship. A researcher may be cited as author of a manuscript without having contributed to the scientific work involved, another who has fulfilled the requirements of authorship is omitted or his/her name does not appear in the order, which would have been appropriate. All these may reflect various degrees of dishonesty and improper scientific conduct. There are even more severe cases, in which scientific research is sponsored, conducted, and sometimes even published by employees of vested parties who prefer their involvement not being disclosed, using proxy authors instead. This form of ghost authorship may sometimes amount to felony. As chair of the Ethics Committee (M.G.) and Dean (E.G.), both at the Sackler Medical School, Tel Aviv University, we report on our insights related to authorship and present 2 representative cases.

This editorial does not deal with gender nor with the genome but with an aspect of scientific misconduct, namely authorship. We found it appropriate to share with the readers of Gender and the Genome some of our experience related to this topic, because of its general importance in scientific publication.

The main goals of scientific publication are to innovate, increase our body of scientific knowledge, promote science, and share scientific achievements with colleagues. There are also additional legitimate reasons for scientific publishing, namely responding to the dictum of “publish or perish.” Publications are required for academic and professional promotion, increasing prestige, and achieving research funding. In this context, authorship of scientific publications and the order in which the names of authors are listed is of great importance, sometimes overshadowing the real purpose of research and not infrequently abused. A search of the PUBMED database with “authorship” as title word, revealed over 1200 articles.

In order to deal with various issues related to scientific publications including authorship, Scientific Journal Editors Committees were setup. ^1,2 The best known of these is the International Committee of Medical Press Editors (ICMJE). ³ In addition, most scientific journals and most academic institutions publish guidelines for authors.

The ICMJE requirements for authorship include all of the following:

Substantial contributions to the conception or design of the work; or the acquisition, analysis, or interpretation of data for the work; and

Drafting the work or revising it critically for important intellectual content; and

Final approval of the version to be published; and

Agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. ³

Technically, it is not always possible to meet all these requirements, especially in large multicenter studies. ⁴ In the past, this was simpler. The 1953 Nobel Prize winning publication of chromosome structure was a one-pager and featured 2 authors, James Watson and Francis Crick. ⁵ In contrast, a 2010 publication lists 2080 authors. ⁶ Of course, not all 2080 authors fulfill the complete list of ICMJE’s requirements, but as a multicenter study, this type of publication includes corresponding authors and a supervisory “ writers group,” who share the responsibility to comply with the directives of the ICMJE.

In general, by fulfilling all 4 of the above requirements an investigator is entitled to be listed as an author of a paper. Unfortunately, the reality may sometimes be slightly different.

Honors, Gifts, Deals, and Ghosts

The various issues related to authorship can be divided into several categories:

The first category refers to authors of a publication who, in spite of having not fulfilled the aforementioned criteria, were included for “honorary” purposes, for example, a junior researcher adding the name of a senior researcher or department head. This might be an attempt to improve the chances of acceptance of the paper or to receive other fringe benefits. There is another variation, in which authorship is provided as a “gift,” meaning that one researcher includes a colleague simply as a friendly gesture. There is the “transaction” version, where 2 investigators decide in advance that one will add the other to the list of authors of an article, and in return his colleague will reward him or her in a similar fashion. Finally, there is “coerced authorship,” where a higher ranking researcher forces a subordinate to include his name among the authors of an article to which he or she had no significant contribution.

All variations in honorary authorship are basically unethical, dilute the position of honest research, are dishonest to readers and editors and degrade research to the level of a commodity. Yet, it is relatively rare to witness a formal complaint. The 2 authors of this article have together accumulated over 50 years as heads of major clinical departments and have never been asked to address a complaint related to honorary authorship.

Another very significant category is “ghost authorship,” in which researchers have taken an active role in a study, fulfilled all or most parts of the ICMJE criteria, but are not listed as authors. A large part of appeals to ethical committees in research deal with this issue.

There may be many reasons for excluding a researcher from the list of authors. It may be that in the principal investigator’s opinion, the contribution of a specific researcher was too small to justify authorship. Alternatively, the paper had undergone several revisions, and the ghost author was involved only at the start of the study. There might be a personal conflict between the researchers. In any event the ghost author who was not listed as a coauthor, or otherwise not acknowledged might feel offended, betrayed, and/or robbed.

There is another version of the ghost author where the researcher has invested heavily in a research project and may even have written the article but is not interested in disclosing his/her involvement in the work. A classic illustration of this is one in which there are clear financial interests in publishing the research, and the investigator hides his involvement from the public. A less known version of ghost authorship is the grey area in which an investigator is listed as an author but does not disclose his financial conflict of interest in the published paper. Under these circumstances, the ghost author essentially serves as an undisclosed proxy for a company. On September 13, 2018, Dr José Baselga, the chief medical officer at the world-renowned Memorial Sloan Kettering Cancer Center in New York was forced to resign because of his failure to disclose that he had received millions of dollars from major health-care companies related to a majority of his research publications. ⁷

In another scandal in 2004, it became clear that the use of rofecoxib (Vioxx) was associated with many severe cardiovascular side effects, including an increase in mortality. ⁸ It was estimated that between 80 000 and 140 000 people were harmed by the medication. Merck & Co, Inc was compelled to remove the medication from the market. By then, annual sales of the drug had reached over US$2.5 billion. It later emerged that although the side effects of the drug were known for about 5 years before the medication was removed from the market, the information had not been reported in more than 95 scientific publications concerning rofecoxib. Most of these publications were financed by Merck and some of the studies were not only funded by the company but were written by company employees whose names were not listed as authors. Only a fraction of these publications reported an affiliation to Merck. These “ghost authors” often even recruit well-known researchers and doctors, sometimes even paying them to be authors on their behalf. The financial damage for Merck, including compensations for the victims amounted to more than US$5 billion. ⁹

Hence, authorship and disclosure of financial ties are not only rights but also duties as part of formal disclosure. If a researcher is not interested in being listed as an author of a publication, his/her contribution should be noted as a part of appropriate disclosure.

The issue of “honorary and ghost authors” involves first-rate journals as well as those less well known. In a study published in 2011, nearly 900 authors of studies published in leading science journals (Annals of Internal Medicine, Journal of the American Medical Association, Lancet. NEJM, PLOS Med, Nature Med) were asked about authorship. ¹⁰ The response rate to the questionnaire was over 70%. Approximately 20% of respondents admitted to honorary or ghost authorship.

Even in the framework of a respected institution such as the Cochrane Collaboration, the phenomenon of “honorary and ghost” is not uncommon. In a study published in 2002, ¹¹ it was discovered that the “honorary authorship” phenomenon existed in 39% and “ghost authorship” in 9% of the reviews.

The following are 2 examples of complaints that were received by the Ethical Committee of the Sackler Faculty of Medicine at Tel Aviv University:

A was involved in a research project of a student, cosigned the submission form and even presented the paper at an international conference. B who was the principal investigator in this research, submitted the manuscript for publication several times and it was rejected. Over time, the connection between A and B was severed due to personal reasons. Two years later, the paper was published and to A’s surprise, his name was not included among the authors. A complained to the Ethics Committee and B responded that he had invested a great deal of effort in the research and that during 2 years of unsuccessful attempts to publish it, A did not show any interest in the fate of the research, did not contribute to its expansion and rewriting, and actually disappeared from the field. The principal investigator (B) extended the study, added a large number of subjects, so that ultimately, according to B it was not the same scope of research that A had participated in at the outset. In addition, B argued, that in any case A’s involvement at the beginning of the process was quite negligible. In light of all this, according to B’s opinion there was no justification for A to be included among the authors of the publication. The Ethics Committee, after examining the data and carefully reviewing the 2 versions of the study, concluded that A’s involvement in the initial path of the study justified the requirement for eventual authorship. The committee came to this conclusion because it found more similarities than differences between the original research and the published work. The committee further stressed that a personal conflict should not prevent the transfer of relevant information between research partners. The committee also pointed out that in any case B should have considered adding A’s name in the acknowledgments section of the study.

The following case also deals with the “ghost author” but more with communication and expectations.

A’s complaint: B concluded a research project but could not publish it because of some additional tests that were required. B requested the use of A’s laboratory to complete these tests. A made available for B a system to carry out the aforementioned tests, which required allocation of resources and personnel. The study was published, but to A’s surprise, his name was not included among the authors. As a result, A filed a complaint with the ethics committee and at the same time with the scientific journal that had published the research. B’s response: the study in all of its stages was concluded prior to the tests in question, and therefore A did not meet the ICMJE criteria for authorship. Furthermore, B assumed that the conduction of the tests was within the framework of a fee for service arrangement and he had transferred to A the amount required for the conduction of the tests. The ethics committee reached the following conclusions: There was an obvious communication failure and unaligned set of expectations between the 2 researchers. The 2 laboratory directors did not conduct even 1 direct meeting in order to define the role of each. The payment transferred by B to A was not a fee for service but only covered consumables. The committee recommended (and the parties accepted the recommendation) that B reimburse A for half the cost of the salary of the doctoral student who performed the tests. B offered to include A’s name in the acknowledgments section of the paper. A withdrew his complaint to the journal.

The committee also addressed the issue of fee for service. The committee stressed that the conducting of laboratory tests for a research project does not automatically grant the right to authorship. In order for a researcher who does laboratory work to be considered as entitling to authorship, there is also a necessity to fulfill additional requirements as described above. All this must be coordinated before the tests are performed. Moreover, fee for service cannot be an alternative to the status of an author. Otherwise it could be concluded that there might be options to choose between monetary compensation or authorship.

What Can We Learn From These Cases?

Formal and transparent planning of authorship should be an important preliminary step in collaboration of scientific work. This includes assignment of tasks regarding the research project, defining expectations, order of authorship listing or mention in the acknowledgment section. All this should be appropriately documented. For laboratory/clinician collaboration, laboratory tests that are performed as part of routine diagnosis and treatment should be distinguished from tests performed specifically for a research project. Here, too, it is desirable to produce appropriate documentation at the onset of the project. As a result on the described case and similar complaints, the committee has designed a structured collaboration form, which is now widely accepted and used at our medical school.

Many misunderstandings, anger, frustration, and aggravation can be prevented by coordinating and clarifying in advance issues regarding the roles of each and every one of the participating researchers, including authorship.