CSE's White Paper on Promoting Integrity in Scientific Journal Publications

3.0 IDENTIFYING RESEARCH MISCONDUCT AND GUIDELINES FOR ACTION

3.5 Correcting the Literature

Because of the breadth of the scientific culture, it is important to note that there is not one recognized method for addressing literature corrections. Of the various scientific disciplines reviewed for this section, the biomedical sciences have had the most experience in addressing literature correction issues. Hence, the information in this section is built largely on the literature correction policies of 2 organizations that have had extensive experience in this area, the National Library of Medicine (NLM) and the International Committee of Medical Journal Editors (ICMJE).

The NLM is the largest medical library in the world, serves millions of researchers through MEDLINE, and develops policies annually in response to issues that surface in the biomedical publishing community. The Uniform Requirements for Manuscripts Submitted to Biomedical Journals, which are endorsed by more than 500 journals, reflect the experiences of editors since 1978 and are updated regularly to address new issues in scientific publication. The guidelines of both these organizations provide a useful framework to the greater scientific research community for addressing the issues related to correcting the literature.

The following sections will examine the state of the art for literature corrections, including definitions, processes, a checklist for editors, and examples for language used for correcting the literature.

3.5.1 Definitions

One of the most confusing aspects associated with literature corrections are the terms journals use when identifying what is being corrected. Different terms are sometimes used interchangeably. For example, the term retraction is not applied by journals uniformly. Some journals will use the term erratum for a retraction and this can lead to confusion for the reader. For the purpose of this document, the definitions used by the NLM will serve as the gold standard for literature correction terminology.

The primary methods used for correcting the literature are errata and retractions.

• Errata. Published changes or emendations to an earlier article, frequently referred to as corrections or corrigenda, are considered by NLM to be errata, regardless of the nature or origin of the error. The NLM does not differentiate between errors that originated in the publication process and errors of logic or methodology.

• Retractions. Retractions identify a citation that was previously published and is now retracted through a formal issuance from the author, publisher, or other authorized agent. The NLM does not differentiate between articles that are retracted because of honest error and those that are retracted because of scientific misconduct or plagiarism. If the notification in the journal is labeled as a retraction or withdrawal, NLM will index it as a retraction.

• Expressions of Concern. This indexing term was introduced by the ICMJE, incorporated into the NLM system in 2004, and has been used on a few occasions.^1,2 The expression of concern is a label that an editor may use to draw attention to possible problems but does not go so far as to retract or correct an article. Examples of this correction format are provided at the end of this section.

3.5.2 Published Guidelines

While a wide variety of journals may be aware of literature correction issues, experiences are not uniform, and established policies and procedures often do not exist. Many disciplines have codes of conduct regarding good publishing practices, but few specifically state how literature corrections will be addressed. Literature corrections are typically handled on a case-by-case basis.

The American Physical Society published the Supplementary Guidelines on Responsibilities of Coauthors and Collaborators (adopted by the APS Council on November 10, 2002) and these guidelines discuss authorship responsibilities associated with maintaining integrity in what is published. In addition to addressing authorship responsibilities, the guidelines state that "all coauthors have an obligation to provide prompt retractions or correction of errors in published works. Any individual unwilling or unable to accept appropriate responsibility for a paper should not be a coauthor."³ While not all authors who publish are members of the American Physical Sciences, anyone who publishes in their journal is held to these standards.

The Society for Neuroscience has been one of the leading professional societies to address literature corrections due to a finding of scientific misconduct. In their 1998 publication, "Responsible Conduct Regarding Scientific Communication," the society outlines the following steps:

If an investigation concerning a published article or abstract determines that the article contains a serious error, then a correction or retraction must be published prominently in the journal or abstract collection in which the original report appeared and contain the full bibliographic reference to the article or abstract. It should also be listed in the contents page and be prominently labeled (eg, erratum, retraction, or apologia).

If the article or abstract was authored by more than one individual and some of those individuals are found to be innocent of misconduct, this should be made clear in the published statement. Any co-authors not found to be guilty of misconduct should be invited to participate in the preparation of the correction or retraction and/or to add an indication of their agreement to the statement. However, such authors should not be permitted to block publication of the statement.⁴

While many disciplines do address the role of the author and the responsibilities associated with ethical publishing practices, many do not address literature corrections. Those fields that do not have an established policy on literature correction issues at this time include botany, chemistry, geophysical science, and veterinary medicine.

3.5.3 The US Public Health Service

The US Public Health Service (PHS) Office of Research Integrity (ORI) has had a wide range of experience with journal editors and authors in reference to publications requiring literature corrections due to findings of scientific misconduct.

The ORI is the office within the PHS that is responsible for addressing scientific misconduct and research integrity related to PHS activities. One of the PHS administrative actions requires the respondent to submit a letter to the editor of the journal in which the article is being corrected due to a finding of scientific misconduct. When a respondent is required to submit a retraction or a correction of an article, the respondent must also send a copy of the retraction or correction letter to the ORI.

To ensure that editors are notified about submitted manuscripts or published articles in their journal that require correction or retraction because of findings of scientific misconduct, the ORI sends the editor a letter with a copy of the Federal Register notice, the ORI report or the voluntary agreement signed by the respondent, and the Departmental Appeals Board decision, if applicable. This notification is sent on publication of the Federal Register notice announcing the PHS findings and administrative actions.⁵

The ORI may request that journals publish corrections or retractions resulting from scientific misconduct cases. Although the ORI does not have authority to require the journal to publish the retraction or correction, it can require the scientist who committed misconduct to submit the request. Besides PHS administrative actions, requests to correct the literature may be initiated by the institution where the misconduct occurred or by a co-author of the questioned paper before the ORI has completed its oversight review. If the request for a retraction is accepted, the editor should publish the retraction as indicated in the Uniform Requirements—meaning it should be labeled as such, appear in a prominent section of the journal, be listed in the table of contents, and include in its heading the title and citation of the original journal article.⁶

3.5.4 The National Science Foundation, Office of Inspector General

The National Science Foundation, Office of The Inspector General (NSF/OIG) addresses allegations of research misconduct in relation to research funded by the NSF. To date, the NSF/OIG has not addressed scientific misconduct cases that have required literature corrections but relies on a grantee's institution to handle literature corrections related to findings of scientific misconduct.⁷

3.5.5 Processes

Literature corrections, whether in the form of errata or retractions, have been made by a variety of "authorized" agents. These agents have included authors, editor(s), publishers, department chairpersons, deans, laboratory directors, and legal counsel. It is important to mention that journals, professional societies, and government bodies have individual policies addressing how literature corrections will be managed. There are very few journals that address literature corrections.^8,9 However, the NLM and Uniform Requirements describe those persons from whom literature corrections will be accepted.

Of the 2 primary forms of literature corrections, "retractions" are sometimes difficult to attain. As indicated by the NLM, retractions are issued for the more serious literature corrections. While they are not issued only when associated with scientific misconduct, they are most easily published when the responsible author(s) submits the request to the editor. History has shown, however, that there are instances in which an author found guilty of scientific misconduct refuses to submit a retraction. Such situations are delicate and vary in difficulty. Because not all journals have policies on how to address literature corrections, editors are sometimes reluctant to publish a retraction without the signature of the author who committed the misconduct. Yet editors should consider their responsibility of reporting accurate information to their readership. The ORI has had an example in which coauthors and a responsible university official submitted a retraction when the original author refused.¹⁰ The latter part of this document provides examples of coauthors submitting retractions when an author guilty of misconduct refused.

As previously discussed, the NLM and the Uniform Requirements developed by the ICMJE are the leaders in issuing guidance and instruction on correcting the literature. The following sections outline the processes used by both.

The NLM uses the following processes for addressing errata and retractions:

Errata. When a publisher, editor, or author has published a labeled, citable erratum to an article that was cited in the MEDLINE database, NLM has amended the citation of the article with a bibliographic reference to the erratum notice to alert users and refer them to the source of the revised information.

The reference to a published erratum notice is in the form of a notification that appears above the article title in the Abstract or Citation formats of PubMed^®. In the MEDLINE format, this information appears in the EIN (Erratum in) field. Although errors may occur in any part of the published article, NLM will add the corrected information to the citation if the erroneous data were incorporated in the original MEDLINE citation. That is, if the error occurred in the article's authorship, title, or abstract, NLM will retain the original citation, if it affects retrieval, but will add the revised data to provide the correct information. If an author's name was misspelled, the corrected name is inserted in the appropriate order and the original misspelling is moved to the end of the author list. Thus, a user who wishes to follow up on all of the authors from the journal issue will be able to retrieve on the misspelled name as well. The notice about the correction will show both the incorrect spelling of the name and the corrected form.

If, however, the error occurred in a portion of the article that is not included in the MEDLINE citation, such as the text, graphs, or tables, only a reference to the published erratum notice will be added to the MEDLINE citation. Brief errata notices are not generally indexed as independent articles. Some substantive articles or letters may, however, comprise published errata. If so, these items will be indexed with the Publication Type PUBLISHED ERRATUM. For those citations having a publication date of 2002 forward, a link will refer back to the citation for the original article. That link appears above the article title in the Abstract or Citation formats of PubMed while in the MEDLINE format the information appears in the EFR (Erratum for) field.

It is NLM's policy that errata will be acknowledged only if they are printed in a citable form; that is, an erratum notice must appear on a numbered page in an issue of the journal that originally published the article. Error notices that are inserted unbound into a journal issue or "tipped" will not be considered part of the permanent bibliographic record. An erratum notice pertaining to a portion of a journal that exists in online format only must be readily discernable in the table of contents of a subsequent issue. NLM does not make changes in the database in response to letters from authors or editors, unless such letters indicate that a substantive published erratum is forthcoming.

Retractions. Articles may be retracted or withdrawn by their authors, academic or institutional sponsor, editor, or publisher because of pervasive error or unsubstantiated or irreproducible data. It is NLM's policy that a retraction will be indexed as a retraction only if it clearly states that the article in question is being retracted or withdrawn, and is signed by an author of the retracted paper or author's legal counsel; by the head of the department, dean, or director of the laboratory where the paper was produced; or by the journal editor. In addition, the retraction must be labeled and published in citable form; that is, the retraction must appear on a numbered page in an issue of the journal that published the retracted article.

NLM does not simply expunge the citation of a retracted article from its indexes or databases but rather links the original to the notice of retraction by adding a Retraction statement after the source of the retracted article on the PubMed Summary display. The bibliographic reference for the retraction notice also appears above the title in the Abstract and Citation formats in PubMed. In the MEDLINE format, it appears in the RIN (Retraction in) field. The MEDLINE record of each retracted article will be given an additional Publication Type of RETRACTED PUBLICATION (PT) as well.¹¹

NLM makes a reciprocal linkage between the retraction statement and the retracted article. That is, the retraction statement is indexed as RETRACTION OF PUBLICATION (PT). The bibliographic reference(s) for the article(s) being retracted appear above the title in the Abstract and Citations formats in PubMed. In the MEDLINE format, they appear in the ROF (Retraction of) field.

Examples of errata and retractions found in MEDLINE can be found at the online NLM fact sheet.

The processes to correcting literature corrections for errata and retractions as addressed in the Uniform Requirements are the following:

Errata. Errors may be noted in published articles that require the publication of a correction or erratum of part of the work. The corrections should appear on a numbered page, be listed in the contents page, include the complete original citation, and link to the original article and vice versa if online. It is conceivable that an error could be so serious as to vitiate the entire body of the work, but this is unlikely and should be handled by editors and authors on an individual basis. Such an error should not be confused with inadequacies exposed by the emergence of new scientific information in the normal course of research. The latter require no corrections or withdrawals.

Retractions. If a fraudulent paper has been published, the journal must print a retraction… The retraction or expression of concern, so labeled, should appear on a numbered page in a prominent section of the print journal as well as in the online version, be listed in the contents page, and include in its heading the title of the original article. It should not simply be a letter to the editor. Ideally, the first author should be the same in the retraction as in the article, although under certain circumstances the editor may accept retractions by other responsible persons. The text of the retraction should explain why the article is being retracted and include a full original citation reference to it.

The validity of previous work by the author of a fraudulent paper cannot be assumed. Editors may ask the author's institution to assure them of the validity of earlier work published in their journals or to retract it. If this is not done, editors may choose to publish an announcement expressing concern that the validity of previously published work is uncertain.¹²

3.5.6 Editor's Checklist

Because literature corrections may occur at different points throughout the publication process, there is not one specific formula that is applicable in all situations. Editors typically address these matters on a case-by-case basis. However, there are some general issues that an editor should consider when addressing a literature correction:

What is the nature of the correction request? Based on the definitions previously outlined, does the literature correction warrant a correction, retraction, or expression of concern? The correction type should be determined by the nature of the correction.

Who makes the request? Ideally, the request should be made by the responsible author(s). However, as noted in an earlier section, there are occasions when a third party may need to make the correction due to differences of opinion or disagreements among authors regarding the responsibility for the retraction. The editor's concern should be correcting the literature so the readership can rely on the information published.

Who writes the correction? Depending on the situation, the literature correction should be made by the author(s) of the paper being corrected. In those situations where there is disagreement, the correction should be written by a responsible institutional official or the editor.

What verbiage should be used for the correction? The readership is best served when the literature correction states what is being corrected. Errata are often typographical errors. Retractions are typically made owing to honest error, or sometimes scientific misconduct. As stated by the ICMJE guidelines, the text of the retraction should explain why the article is being retracted and include the full original citation. Examples of verbiage used are provided at the end of this document.

When should the correction be published? Depending on the situation, an editor should publish the correction as soon as reasonably possible. In the case in which the corrections are the product of a scientific misconduct investigation, this would occur after a finding of scientific misconduct had been made by an institution or an oversight agency, if appropriate.

On the rare occasion in which a paper under review for possible scientific misconduct included a public health concern, it would be prudent for the institution conducting the investigation to notify the journal editor of this public health concern. The decision of when to publish a retraction then rests with the editor.

3.5.7 Examples of Literature Corrections

In addition to policies varying on how to publish literature corrections, so do the actual publications themselves. The following section provides a variety of literature corrections (errata and retractions), along with identification of who submitted the literature correction. The literature corrections are from publicly available sources and vary in presentation to reflect the authenticity and style of the respective journal.

1. J Infect Dis. 2004;190:2059. Erratum submitted by authors.

   In an article in the 1 November 2004 issue of the Journal (Gumbo T, Louie A, Deziel MR, Parsons LM, Salfinger M, Drusano GL. Selection of a moxifloxacin dose that suppresses drug resistance in Mycobacterium tuberculosis, by use of an in vitro pharmacodynamic infection model and mathematical modeling. J Infect Dis. 2004;190:1642-51), a ">" should have preceded "1 mg/L" in the sixth line in the right-hand column of page 1644. The authors regret this omission.

2. Arch Gen Psychiatry. 2006;63:365. Erratum submitted by authors.

   Errors in Text. In the Original Article by Birmaher et al titled "Clinical Course of Children and Adolescents With Bipolar Spectrum Disorders," published in the February issue of the ARCHIVES (2006;63:175-183), errors occurred in the text on pages 176 and 179. On page 176, in the "Methods" section, "Subjects" subsection, fifth paragraph, the third sentence should have read as follows: "Subjects with BP-II had the onset of their mood disorders significantly later and had significantly lower rates of comorbid attention-deficit/hyperactivity disorder than subjects with BP-I and BP-NOS (P≤.05)." On page 179, under "Weekly Mood Symptomatic Status by BP Subtype," first paragraph, the second sentence should have read as follows: "Within the syndromal symptoms, subjects with BP-I spent significantly more weeks with syndromal mania and mixed symptoms than those with BP-NOS, and subjects with BP-II spent significantly more time with depressive symptoms than those with BP-I and BP-NOS (all comparisons, P≤.001)."

3. Science. 2004;306:54. Retraction submitted by all authors.

   In the Report "Synaptic changes in layer 2/3 underlying map plasticity of developing barrel cortex" (Petersen CC, Brecht M, Hahn TT, Sakmann B. Science. 2004;304(5671):739-742), we concluded that functional and anatomical changes in layer 2/3 underlie different forms of cortical map plasticity. It was pointed out to us by a reader that the anatomical analysis contains errors. Although these errors did not affect the main conclusions, we re-analyzed the data set. Re-analysis confirmed that whisker stimulation evokes a cortical response, which spreads preferentially to neighboring, nondeprived cortical columns as originally reported. However, the reported difference between the axonal fields in control and deprived animals was not statistically significant. Further, the deprivation-induced decrease in unitary EPSP amplitude was also not statistically significant. Thus, major conclusions of the Report are no longer supported, and we retract the Report. We apologize for any confusion that we may have caused to the readers of Science.

4. Proc Natl Acad Sci U S A. 2000;97:1949. Retraction submitted by co-authors but not the author guilty of scientific misconduct.

   For the article "Sodium channels in the cytoplasm of Schwann cells" by J. M. Ritchie, J. A. Black, S. G. Waxman, and K. J. Angelides, which appeared in number 23, December 3, 1990, of Proc Natl Acad Sci U S A (87, 9290-9294), the undersigned authors would like to note the following: "This paper included immunocytochemical studies using antibody 7493. We interpreted immunostaining with antibody 7493 as providing information about sodium channel localization based on an immunological characterization of antibody 7493 carried out in the laboratory of K. J. Angelides. As reported in the Federal Register on March 12, 1999, based on the report of an investigation by the Baylor College of Medicine and on information obtained by the National Institutes of Health Office of Research Integrity (ORI) during its oversight review into allegations of scientific misconduct by Angelides, ORI, on March 10, 1997, found that Angelides falsified the description of the data in the corresponding text and legend of Fig. 1 of this paper and that his conduct constituted scientific misconduct. The Appeals Board of the Department of Health and Human Services (DAB) issued a decision on February 5, 1999, in which it affirmed the findings of ORI. Given the allegations of irregularity in the immunological characterization of antibody 7493 and the findings that ORI and DAB have made, we cannot stand behind the interpretation of results using this antibody. We therefore retract the immunocytochemical and immunoultrastructural results presented in this paper." (J. M. Ritchie, J. A. Black, S. G. Waxman)

5. Proc Natl Acad Sci U S A. 1997;94:12732. Retraction submitted by co-author but not the author guilty of scientific misconduct.

   An author (Hans-Jürgen Gruss) of the article "Tumor necrosis factor receptor-associated factor (TRAF)-1, TRAF-2, and TRAF-3 interact in vivo with the CD30 cytoplasmic domain; TRAF-2 mediates CD30-induced nuclear factor kappa B activation" by Stéphane Ansieau, Inka Scheffrahn, George Mosialos, Heike Brand, Justus Duyster, Kenneth Kaye, Josephine Harada, Bill Dougall, Gabi Hübinger, Elliott Kieff, Friedhelm Herrmann, Achim Leutz, and Hans-Jürgen Gruss, which appeared in number 24, November 26, 1996, of Proc Natl Acad Sci U S A (93, 14053-14058), has admitted scientific misconduct in misrepresenting data including Figs. 2C and 3. Because the experiments of Professor Gruss are a major part of this publication, I request that the paper be withdrawn. (Elliott Kieff)

6. Proc Natl Acad Sci U S A. Published online before print October 6, 2004, doi: 10.1073/pnas.0406725101. 2004;101:15271. Retraction submitted by editors.

   For the article "Prevention of renovascular and cardiac pathophysiological changes in hypertension by angiotensin II type 1 receptor antisense gene therapy," by Jeffrey R. Martens, Phyllis Y. Reaves, Di Lu, Michael J. Katovich, Kathleen H. Berecek, Sanford P. Bishop, Mohan K. Raizada, and Craig H. Gelband, which appeared in issue 5, March 3, 1998, of Proc Natl Acad Sci U S A (95, 2664-2669), after an investigation by the Office of Research Integrity (ORI), Craig H. Gelband admitted to falsification of data, including Fig. 4 A and B. ORI determined that Dr. Gelband is solely responsible for the falsification. The editors, therefore, hereby retract the paper.

7. BMJ. 1998;316:1700. Retraction submitted by the editor.

   The BMJ is retracting the paper by MH Williams and C Bowie (BMJ. 1993;306:95-98) at the request of Dr Bowie. The General Medical Council found Dr Williams guilty of professional misconduct in February 1998 on charges which included research fraud. Dr Williams was responsible for the data collection of the original interview and examination survey in 1989 and the follow up telephone survey in 1990. Dr Bowie has been unable to verify that the data collection was carried out in an honest way. He did not scrutinize the data sheets at the time of the surveys; the data sheets of both surveys have been destroyed; and none of the 18 people still alive in Somerset and contacted by telephone six years later could remember the telephone interview.

8. Gut.
2004;53:774. Retraction submitted by the editor.

Due to an administrative error, one article has been published on two occasions. The journal would like to retract the paper by Lindsay et al in the July issue (Gut. 2003;52:981-987) as it is a replicated version of a paper by the same authors in the March issue (Gut. 2003;52:363-369). The journal apologizes for this error.

9. Gut. 2001;48:286. Retraction submitted by the editor.

   Gut is retracting the paper by AK Banerjee and TJ Peters, "Experimental non-steroidal anti-inflammatory drug induced enteropathy in the rat—similarities to inflammatory bowel disease and effect of thromboxane synthetase inhibitors" (Gut. 1990;31:1358-1364) and the abstract AK Banerjee, R Sherwood, JA Rennie and TJ Peters, "Sulphasalazine reduces indomethacin induced changes in small intestinal permeability in man" (Gut. 1990;31:A593) at the request of Dr Banerjee. At the end of November 2000, the General Medical Council found Dr Banerjee guilty of serious professional misconduct and suspended him for 12 months. Both articles were deemed to contain information which was deliberately falsified.

10. Biotechnol Adv. 2004;22:619. Retraction submitted by the editor.

    The article "Biotransformation of drugs by microbial cultures for predicting mammalian drug metabolism" (Srisilam K, Veeresham C. Biotechnol Adv. 2003;21:3-39) has been retracted at the request of the editors because the authors had infringed the normal professional ethical codes by plagiarizing another publication: "Microbial models for drug metabolism" (Adv Biochem Eng Biotechnol. 1999;63:69–218).

11. Virus Res. 2004;106:83. Retraction submitted by the publisher.

    Retraction of "Nuclear factor kappa B (NF?B) dependent modulation of Epstein-Barr virus latent membrane protein 1 (LMP1) in epidermal growth factor receptor (EGFR) promotor activity" (Tao YG, Tan YN, Liu YP, Song X, Zeng L, Gu HH, Tang M, Li W, Yi W, Cao Y. Virus Res. 2004;104:61–70.) The publisher would like to announce that this paper has been retracted. A paper by the same group of authors containing essentially the same data and conclusions was published a short time earlier (Cell Signal. 2004;16:781–790). The authors have agreed to withdraw their paper from Virus Research.

12. J Med Genet. 2004;41:813. Retraction submitted by authors.

    The authors of the following manuscript (Ninis VN, Kýlýnc MO, Kandemir M, Daõlý E, Tolun A. High Frequency of T9 and CFTR Mutations in Children with Idiopathic Bronchiectasis. J Med Genet. 2003;40:530-535) are retracting it because the polythymidine track genotype data are not correct. Recently the authors repeated the genotyping or 17 of the subjects to check whether the reported genotypes were correct and found out that they were not. At the time of submission of the manuscript, the authors were very confident of the data, since they had employed two independent methods for the genotyping of all subjects. However, subsequently the authors were prompted to recheck the results and have been unable to confirm them. The authors regret that we did not find out prior to publication.

13. J Clin Invest. 2003;112:1265. Retraction submitted by investigative panel.

    The following manuscripts were part of an investigation in Germany.

    Herrmann F, Oster W, Meuer SC Lindemann A, Mertelsmann RH. Interleukin 1 stimulates T lymphocytes to produce granulocyte-monocyte colony-stimulating factor. J Clin Invest. 1988;81:1415-1418.

    Lindemann A, Riedel D, Oster W, Ziegler-Heitbrock HW, Mertelsmann R, Herrmann F. Granulocyte-macrophage colony-stimulating factor induces cytokine secretion by human polymorphonuclear leukocytes. J Clin Invest. 1989;83:1308-1312.

    Oster W, Cicco NA, Klein H, Hirano T, Kishimoto T, Lindemann A, Mertelsmann RH, Herrmann F. Participation of the cytokines interleukin 6, tumor necrosis factor-alpha, and interleukin 1-beta secreted by acute myelogenous leukemia blasts in autocrine and paracrine leukemia growth control. J Clin Invest. 1989;84:451-457.

    Nehls MC, Brenner DA, Gruss H-J, Dierbach H, Mertelsmann R, Herrmann F. Mithramycin selectively inhibits collagen-α 1(I) gene expression in human fibroblast. J Clin Invest. 1993;92:2916-2921.

    These manuscripts were evaluated as part of the Task Force Friedhelm Hermann, a group that investigated the findings published from the lab of Friedhelm Hermann for the Deutsche Forschungsgemeinschaft. The independent committee reviewed concerns related to the validity of the data associated with the above papers. As a result of the committee's findings, we are issuing a retraction of these papers. However, not all contributions by all authors of the papers were found to be fraudulent, and some authors have stated that their experimental contributions were legitimate.

14. Nat Biotechnol. 2004;22:601. Retraction submitted by all authors.

    Retraction: Zeytun A, Jeromin A, Scalettar BA, Waldo GS, Bradbury ARM. Fluorobodies combine GFP fluorescence with the binding characteristics of antibodies. Nat Biotechnol. 2003;21(12):1473-1479. In this article, we concluded that inserting HCDR3 sequences derived from antibodies into a particular stable form of GFP created intrinsically fluorescent affinity reagents, which we termed "fluorobodies." We have recently realized that the strategy used to generate fluorobodies was flawed: the adaptor sequences reported contained additional nucleotides that introduced translational frameshifts at each HCDR3 insertion site. We expect this to have resulted in the creation of tripartite nonfluorescent "GFP-HCDR3 fragments" with the following structures: N-terminal in-frame GFP peptide/HCDR3 sequence/out-of-frame C-terminal GFP peptide.

    Reexamination of the primary bacterial stocks of fluorobodies has revealed that these were not monoclonal, as originally assumed, but contained plasmid mixtures encoding both unmodified GFP and GFP-HCDR3 fragments. Segregation of these plasmids revealed that only colonies containing unmodified GFP genes were green. Following these discoveries, we have subsequently carried out immunofluorescence experiments with the remaining anti-tubulin fluorobody used in the published paper and obtained identical (microtubule-like) staining patterns, whereas GFP or anti-tubulin fluorobody prepared from frozen bacterial stocks gave nonspecific staining. With this exception, no original fluorobody protein preparations were available, and attempts to reproduce other results reported in the paper with fluorobodies expressed from frozen bacterial stocks were unsuccessful. We have reason to believe these stocks may have been compromised, and that the similarity of these stored stocks to the original stocks is questionable.

    To determine whether correctly assembled fluorobodies would be functional, we have subsequently generated libraries of HCDR3s inserted at the single loop sites described in the paper, as well as into three (1-3) and four (1-4) loops simultaneously, within the context of a phage display vector. Only one of the single loop libraries (loop 3: in which 30% of clones had fluorescence greater than 5% of the GFP fluorescence, with a maximum of 12%) contained significant numbers of fluorescent clones; the rest were essentially nonfluorescent.

    With the exception of the immunofluorescence data (produced by A.J. and B.A.S. using material prepared by A.Z.), all experiments were carried out by A.Z. in A.R.M.B.'s laboratory. We are presently unable to explain the biological activity described in the original publication. The possibility that this arose from noncovalent association of GFP with GFP-HCDR3 fragments in bacteria containing multiple plasmids is presently under investigation. In light of these findings, all authors agree that this study has not demonstrated the creation of fluorobodies as described, and consequently wish to unanimously retract the paper.

15. Plant Cell. 2004;16:785. Retraction submitted by all authors.

    Jason W. Lilly, Jude E. Maul, and David B. Stern. The Chlamydomonas reinhardtii Organellar Genomes Respond Transcriptionally and Post-Transcriptionally to Abiotic Stimuli. Plant Cell. 2002;14:2681-2706. The authors of this article have requested that its publication be retracted from The Plant Cell. This follows a finding of the Boyce Thompson Institute for Plant Research that Dr Jason Lilly engaged in scientific misconduct, having falsified microarray data found in Figure 4 and the supplementary data set. The authors have further determined that a significant number of clones on the microarray were incorrectly annotated, and they have been unable to reproduce the increased accumulation of certain chloroplast mRNAs in response to sulfur deprivation. The authors wish to emphasize that Dr Lilly was found to be solely responsible for the scientific misconduct and misleading data associated with this publication. They deeply regret any inconvenience resulting from the publication of his data.

16. N Engl J Med. 2003;348:2137. Expression of concern submitted by editors.

    In the issue of January 31, 2002, we published a study by Helmut Schiffl, MD, Susanne M. Lang, MD, and Rainald Fischer, MD (Daily hemodialysis and the outcome of acute renal failure. N Engl J Med. 2002;346:305-310). It has come to our attention, through communication with Klaus Peter, Dean of the Medical Faculty at Ludwig Maximilians University in Munich, Germany, that there is an ongoing investigation into potential scientific misconduct in the performance of this study. We will inform our readers of the outcome of this investigation when it is complete.

17. Proc Natl Acad Sci U S A. 2003;100:11816. Expression of concern submitted by editors.

    Editorial Expression of Concern: The editors express a note of concern regarding the article "Preferential repair of ionizing radiation-induced damage in the transcribed strand of an active human gene is defective in Cockayne syndrome," by Steven A. Leadon and Priscilla K. Cooper, which appeared in issue 22, November 15, 1993, of Proc Natl Acad Sci U S A (90, 10499-10503).

    An ad hoc committee at the University of North Carolina at Chapel Hill (UNC) has concluded that the results published by Dr Steven A. Leadon, former Professor of Radiation Oncology in the School of Medicine at UNC, which are based on his monoclonal antibody assays for transcription-coupled repair (TCR), should not be relied on unless independent verification exists.

    After reviewing laboratory notebooks, the investigating committee could not confirm that equal amounts of DNA were loaded onto gel lanes that were then assayed for TCR. The committee concluded that the reported preferential repair of the transcribed DNA strand was not supported by available photographs of ethidium bromide—stained gels. The committee further concluded that Dr Leadon was solely responsible, at least for the last 7 years, for the step of the assay that determined the loading of the gel lanes. In addition, in the opinion of the UNC committee, this biased loading was deliberate and done without the knowledge of other scientists in his laboratory or his collaborators.

    As a consequence of this investigation, the UNC committee requested that PNAS evaluate the results of the above-cited paper, which depends critically, but not exclusively, on Dr Leadon's TCR assay.

    We have investigated the matter and are concerned about the validity of the results. We know of no independent verification of the data in the published figures. We therefore think it reasonable for the scientific community to view with extreme caution the results of these assays in the PNAS article. The editors emphasize that our skepticism does not extend to the validity of TCR, which has been amply corroborated by other experiments.

    Dr Leadon does not concur with this assessment and note of concern. Although Dr Cooper cannot of her own knowledge dispute the stated concern with the TCR data, she attests that the conclusions from the paper are valid, based on subsequent work in several laboratories, including her own.

(Authorship: Mary Scheetz took the lead in writing this section of the white paper on behalf of the CSE Policy Committee. Members of the 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 September 13, 2006.)

1. S. Kotzin, Chief, Indexing, MEDLINE; written communication, December 2004.
2. International Committee of Medical Journal Editors. Uniform Requirements for Manuscripts Submitted to Biomedical Journals. Available at: http://www.icmje.org/#correct. Accessed April 26, 2006.
3. American Physical Society. Supplementary guidelines on responsibilities of coauthors and collaborators. Available at: http://www.aps.org/statements/02_2.cfm#supplementary_guidellines1. Accessed April 26, 2006.
4. Society for Neuroscience. Dealing with possible scientific misconduct. Available at: http://web.sfn.org/index.cfm?pagename=responsibleConduct_dealingWithPossibleScientificMisconduct. Accessed April 26, 2006.
5. The ORI has adopted a target timeline of 480 days for completing misconduct cases that involve research supported by the PHS. The timeline begins with the initiation of an institutional inquiry and concludes with review by the Assistant Secretary for Health. Cases that are appealed to the Departmental Appeals Board (DAB) or investigated by the Office of the Inspector General (OIG) are not included, because the DAB regulation establishes 9 months as a goal for completion of a hearing and the OIG is independent from Departmental supervision. Extensions are granted for reasonable cause. The general timeline can be found at: http://ori.hhs.gov/misconduct/inquiry_issues.shtml. Accessed April 26, 2006.
6. Office of Public Health and Science. Managing allegations of scientific misconduct: a guidance document for editors. Available at: http://ori.hhs.gov/documents/masm_2000.pdf. Accessed April 26, 2006.
7. J. Kroll, Head of Administrative Investigation, NSF/OIG; written communication, January 2005.
8. Scheetz MD. Coming full circle: can misconduct be prevented? Presented at: The Journal's Role in Scientific Misconduct: An Educational Retreat. Leesburg, Va; November 9, 2003.
9. Scheetz MD. Promoting integrity through instructions to authors: a preliminary analysis. In: Steneck NH, Scheetz MD, eds. Proceedings of the 1st ORI Research Conference on Research Integrity. Available at: http://69.59.142.46/documents/instructions_authors.pdf. Accessed April 26, 2006.
10. An investigation conducted by the University of California, San Francisco, found that an author falsified data in a publication on AIDS research. According to the investigation, he selectively suppressed data that did not support his hypothesis and reported consistently positive data even though only 1 of his 4 experiments had produced positive results. The falsified data were then used as the basis for a grant application to the National Institutes of Health. The ORI concurred in the university's finding. The researcher executed a "voluntary exclusion and settlement agreement" with PHS in which he agreed not to apply for federal grant or contract funds and would not serve on PHS advisory committees, boards, or peer review groups for 3 years. The publication was retracted. When the author refused to agree to a retraction, the New England Journal of Medicine published the retraction without his signature but with the signatures of the rest of the coauthors and of the assistant vice chancellor of the university. Case study presented at: The Journal's Role in Scientific Misconduct: An Educational Retreat. Leesburg, Va; November 9, 2003.
11. National Library of Medicine. Errata, retraction, duplicate publication, comment, update and patient summary policy for MEDLINE. Available at: http://www.nlm.nih.gov/pubs/factsheets/errata.html. Accessed April 26, 2006.
12. International Committee of Medical Journal Editors. Uniform Requirements for Manuscripts Submitted to Biomedical Journals. Available at: http://www.icmje.org/#correct. Accessed April 26, 2006.

Return to Top

Return to the Promoting Integrity in Scientific Journals table of contents page.