Hide Notes 

Slide 1: Accountability in Research, 13:247–258, 2006 Copyright © Taylor & Francis Group, LLC ISSN: 0898-9621 print / 1545-5815 online DOI: 10.1080/08989620600848199 Accountability in Research: Vol. 13, No. 3, June 2006: pp. 1–15 1545-5815 0898-9621 GACR Research RESEARCH MISCONDUCT POLICIES OF HIGH IMPACT BIOMEDICAL JOURNALS BARBARA K. REDMAN Department of Medical Ethics, University of Pennsylvania, Philadelphia, PA, USA, and Wayne State University, College of Nursing, Detroit, MI, USA JON F. MERZ Department of Medical Ethics, University of Pennsylvania, Philadelphia, PA, USA B.K. Redman and J.F. Merz Policies on Research Misconduct Several national and international organizations have recommended policies regarding journal responsibilities about research misconduct in submitted or published manuscripts. A search of Web sites of the fifty highest impact journals in a cluster of biomedical fields and a limited survey of their editors shows that few journals have formally adopted standards for dealing with questions of research misconduct. Publicly available policies may have a deterrent effect and can prevent arbitrariness in handling cases. Keywords: research, misconduct, journal, policies, allegations, manuscripts Infamous cases of scientific misconduct over the last several decades involving John Darsee (Knox, 1983), Robert Slutsky (Locke, 1986), Roger Poisson (Lowry, 1994), and Jan Hendrik Schon (Service, 2002) have led to high profile retractions recorded in Medline over the 1983–2002 period. Nonetheless, sanctions for discovered scientific misconduct are not a particularly good measure of the frequency of misconduct, and retractions are perhaps the tip of the iceberg (Woolf, 1981). Journals are a critical element in maintaining scientific integrity, because journal Barbara K. Redman designed the study, collected and analyzed data, and wrote and edited the article. Jon F. Merz designed the study and wrote and edited the study. Disclosure: There are no financial or personal conflicts of interest associated with this article. Address correspondence to Barbara K. Redman, Department of Medical Ethics, University of Pennsylvania, 3401 Market Street, Suite 320, Philadelphia, PA 19104-3308, USA. E-mail: ae9080@wayne.edu 247
Slide 2: 248 B.K. Redman and J.F. Merz articles are at once the scientist’s primary means for communicating with peers, the primary mechanism for peer review, and, when misconduct occurs, a source of potential error in science. Indeed, a recent study has shown that retracted studies continue to be cited (Budd et al., 1999). It is thus critical that journals have policies for dealing with allegations of misconduct. Both the International Commission of Medical Journal Editors (ICMJE) and the Committee on Publication Ethics (COPE) (U.K.) provide consensus standards for the responsibility of journal editors if research misconduct is suspected or confirmed in submitted or published manuscripts. In 2000, the U.S. Office of Research Integrity (ORI) provided guidance for editors managing allegations of scientific misconduct (see Table 1), particularly about procedures for handling suspect manuscripts, responding to ORI request for assistance, and correcting the literature. The purposes of this study are to examine whether highimpact biomedical journals have adopted misconduct policies and to describe the policies that are in place. Methods The Journal Citation Report (Science Edition) is a bibliometric analysis of science journals in the Information Sciences Institutes database. Impact factor is defined as the average number of times articles in a specific journal appearing in a two year time period (2001–2002) were cited in the Journal Citation Report the following year (2003). It was created to help evaluate a journal’s relative impact on the field. Table 2 identifies the fifty highest impact biomedical journals and the fields from which they were drawn.(Journal Citation Reports, 2004). Each journal Web site or the first issue of 2005 were searched by one of us (B. K. Redman) in February of 2005 for Instructions for Authors and other documents that described journal policies toward research misconduct (Web site addresses may be found in Table 3). These policies were then content-analyzed to describe the definition of research misconduct and the process by which an allegation was addressed. Editors of these journals were contacted by e-mail with telephone follow-up to verify the existence of a policy, why one was adopted or whether adoption was being considered, and the definition of research misconduct; the
Slide 3: Policies on Research Misconduct TABLE 1 Suggested journal policies for research misconduct 249 “If substantial doubts arise about the honesty or integrity of work either submitted or published, it is the editor’s responsibility to ensure that the question is appropriately pursued, usually by the author’s sponsoring institution. However, it is not ordinarily the task of editors to conduct a full investigation or to make a determination; that responsibility lies with the institution where the work was done or with the funding agency. . . . If this method of investigation does not result in a satisfactory conclusion, the editor may choose to conduct his or her own investigation . . . ” International Committee of Medical Journal Editors: Uniform Requirements for Manuscripts Submitted to Biomedical Journals: Writing and Editing for Biomedical Publication, updated October 2004. “ . . . Editors should not simply reject papers that raise questions of misconduct. They are ethically obligated to pursue the case . . . COPE is always willing to advise, but for legal reasons, can only advise on anonymised cases. It is for the editor to decide what action to take . . . Editors must take all allegations and suspicions of misconduct seriously, but they must recognize that they do not usually have either the legal legitimacy or the means to conduct investigations into serious cases. The editor must decide when to alert the employers of the accused author(s). Some evidence is required, but if employers have a process for investigating accusations—as they are increasingly required to do—then editors do not need to assemble a complete case. Indeed, it may be ethically unsound for editors to do so, because such action usually means consulting experts, so spreading abroad serious questions about the author(s). If editors are presented with convincing evidence—perhaps by reviewers—of serious misconduct, they should immediately pass this on to the employers, notifying the author(s) that they are doing so. If accusations of serious misconduct are not accompanied by convincing evidence, then editors should confidentially seek expert advice. If the experts raise serious questions about the research, then editors should notify the employers . . . ” Committee on Publication Ethics (COPE): Guidelines on Good Publication Practice, 1999. “Instructions to authors should state that the authors, by submitting a manuscript to the journal, will abide by the journal’s policy and procedures for notifying the author’s institution or ORI. This notification should also state that authors agree to cooperate with an institution or ORI in investigating an allegation of scientific misconduct involving their manuscript or article. Developing procedures for handling suspect manuscripts will guide the editorial staff and reviewers.” Office of Research Integrity: Managing Allegations of Scientific Misconduct; A Guidance Document for Editors, 2000.
Slide 4: TABLE 2 50 highest impact biomedical journals* Nature Reviews Immunology Nature Reviews Neuroscience Nature Reviews of Cancer Nature Reviews of Molecular Cell Biology Neuron New England Journal of Medicine Pharmacological Reviews Physiological Review Progress in Neurobiology Science Trends in Biochemical Science Trends in Cell Biology Trends in Ecology and Evolution Trends in Immunology Trends in Neuroscience Trends in Pharmacological Sciences 250 Immunity Journal of American Medical Association Journal of Clinical Investigation Journal of Experimental Medicine Journal of the National Cancer Institute Lancet Microbiology and Molecular Biology Reviews Molecular Cell Nature Nature Biotechnology Nature Cell Biology Nature Genetics Nature Immunology Nature Medicine Nature Neuroscience Nature Review Drug Discovery Nature Reviews Genetics Annals of Internal Medicine Annual Review Biophysiology and Biomedicine Annual Review Cell and Developmental Biology Annual Review of Biochemistry Annual Review of Immunology Annual Review of Neuroscience Annual Review of Pharmacology Annual Review Physiology CA – Cancer Journal for Clinicians Cancer Cell Cell Current Opinion in Cell Biology Current opinion in Genetics and Development Developmental Cell Endrocrine Reviews Gastroenterology Genes and Development *Journals with the highest impact factors in the 2003 Journal Citation Report Science Edition were selected from the following medical categories: Allergy; Anatomy & Morphology; Anesthesiology; Behavioral Sciences; Biochemistry & Molecular Biology; Biology; Biophysics; Biotechnology & Applied Microbiology; Cardiac & Cardiovascular Systems; Cell Biology; Clinical Neurology; Critical Care Medicine; Dentistry, Oral Surgery & Medicine; Dermatology; Developmental Biology; Emergency Medicine; Endocrinology & Metabolism; Gastroenterology & Hepatology; Genetics & Heredity; Geriatrics & Gerontology; Health Care Sciences & Services; Hematology; Immunology; Infectious Diseases; Integrative & Complementary Medicine; Medical Ethics; Medical Informatics; Medical Laboratory Technology; Medicine, General & Internal; Medicine, Research & Experimental; Microbiology; Multidisciplinary Sciences; Neuroimaging; Neurosciences; Nursing; Nutrition & Dietetics; Obstetrics & Gynecology; Oncology; Ophthalmology; Orthopedics; Otorhinolaryngology; Pediatrics; Peripheral Vascular Disease; Pharmacology & Pharmacy; Physiology; Psychiatry; Psychology; Public, Environmental & Occupational Health; Radiology, Nuclear Medicine & Medical Imaging; Rehabilitation; Reproductive Biology; Respiratory System; Rheumatology; Sport Sciences; Substance Abuse; Surgery; Toxicology; Transplantation; Tropical Medicine; Urology & Nephrology; Virology; Zoology.
Slide 5: TABLE 3 Internet addresses of the 50 highest impact biomedical journals* where information was obtained for study Immunity http://www.immunity.com/ misc/page?page=authors Journal of American Medical Association http://jama.ama-assn.org Nature Reviews Immunology http://www.nature.com/ni/authors/ index.html Nature Reviews Neuroscience http://www.nature.com/nrn/index.html Nature Reviews of Cancer http://www.nature.com/nrc/index.html Nature Reviews of Molecular Cell Biology http://www.nature.com/nrm/index.html Neuron http://www.neuron.org/misc/ page?page=authors New England Journal of Medicine http://authors.nejm.org/ Pharmacological Reviews http://pharmrev.aspetjournals.org/misc/ ifora.shtml (Continued) Annals of Internal Medicine http://www.annals.org/shared/ edit_staff.shtml Annual review Biophysiology and Biomedicine http://www.annualreviews.org/ Annual Review Cell and Developmental biology http://www.annualreviews.org/ Annual Review of Biochemistry http://www.annualreviews.org/ Annual Review of Immunology http://www.annualreviews.org/ Journal of Clinical Investigation http://www.jci.org/misc/contact.shtml Journal of Experimental Medicine http://www.jem.org/ Journal of the National Cancer Institute http://jncicancerspectrum.oxfordjournals.org/ Lancet http://www.thelancet.com/ Microbiology and Molecular Biology Reviews http://mmbr.asm.org/misc/ ifora.shtml Annual Review of Neuroscience http://www.annualreviews.org/ Annual Review of Pharmacology http://www.annualreviews.org/ 251
Slide 6: 252 Molecular Cell http://www. molecule.org/ Nature http://www.nature.com/ nature/authors/index.html Physiological Review http://www. the-aps.org/publications/journals/ pub_quick.htm Progress in Neurobiology http:// www.elsevier.com/wps/find/ journaldescription.cws_home/412/ authorinstructions Science http://www.sciencemag.org/ feature/contribinfo/home.shtml Nature Biotechnology http://www.nature.com/nbt/ authors/ index.html Nature Cell Biology http:// www.nature.com/ncb/ authors/index.html Nature Genetics http:// www.nature.com/ng/authors/ index.html Nature Immunology http://www.nature.com/ni/ authors/index.html Nature Medicine http://www.nature.com/nm/ authors/index.html Trends in Biochemical Science http:// www.elsevier.com/wps/find/ journaldescription.cws_home/405916/ authorinstructions Trends in Cell Biology http://www.elsevier.com/wps/find/ journaldescription.cws_home/422552/ authorinstructions Trends in Ecology and Evolution http://www.elsevier.com/wps/find/ journaldescription.cws_home/30339/ authorinstructions Trends in Immunology http:// www.elsevier.com/wps/find/ journaldescription.cws_home/405914/ authorinstructions TABLE 3 (Continued) Annual Review Physiology http://www.annualreviews.org/ CA – Cancer Journal for Clinicians http://caonline.amcancersoc.org/misc/ ifora.shtml Cancer Cell http://www.cancercell.org/ misc/page?page=authors Cell http://www.cell.com/misc/ page?page=authors Current opinion in cell biology http://www.current-opinion.com/jcel/ about.htm?jcode=jcel Current opinion in Genetics and Development http://www.current-opinion.com/ jgen/about.htm Developmental Cell http://www. developmentalcell.com/misc/ page?page=authors
Slide 7: Endrocrine Reviews http://edrv. endojournals.org/misc/ifora.shtml Nature Neuroscience http:// www.nature.com/neuro/authors/ index.html Nature Review Drug Discovery http://mts-nrdd.nature.com/cgi-bin/ main.plex?form_type=display_auth_ instructions Nature Reviews Genetics http://mts-nrg.nature.com/cgi-bin/ main.plex?form_type=display_auth_ instructions Gastroenterology http://www2.gastrojournal.org/scripts/om.dll/serve?web=/info/ wbs-info.htm&id=gast Trends in Neuroscience http:// www.elsevier.com/wps/find/ journaldescription.cws_home/405919/ authorinstructions Trends in Pharmacological Sciences http://www.elsevier.com/wps/find/ journaldescription.cws_home/405920/ authorinstructions Genes and Development http:// www.genesdev.org/misc/ifora.shtml *Journals with the highest impact factors in the 2003 Journal Citation Report Science Edition were selected from the following medical categories: Allergy; Anatomy & Morphology; Anesthesiology; Behavioral Sciences; Biochemistry & Molecular Biology; Biology; Biophysics; Biotechnology & Applied Microbiology; Cardiac & Cardiovascular Systems; Cell Biology; Clinical Neurology; Critical Care Medicine; Dentistry, Oral Surgery & Medicine; Dermatology; Developmental Biology; Emergency Medicine; Endocrinology & Metabolism; Gastroenterology & Hepatology; Genetics & Heredity; Geriatrics & Gerontology; Health Care Sciences & Services; Hematology; Immunology; Infectious Diseases; Integrative & Complementary Medicine; Medical Ethics; Medical Informatics; Medical Laboratory Technology; Medicine, General & Internal; Medicine, Research & Experimental; Microbiology; Multidisciplinary Sciences; Neuroimaging; Neurosciences; Nursing; Nutrition & Dietetics; Obstetrics & Gynecology; Oncology; Ophthalmology; Orthopedics; Otorhinolaryngology; Pediatrics; Peripheral Vascular Disease; Pharmacology & Pharmacy; Physiology; Psychiatry; Psychology; Public, Environmental & Occupational Health; Radiology, Nuclear Medicine & Medical Imaging; Rehabilitation; Reproductive Biology; Respiratory System; Rheumatology; Sport Sciences; Substance Abuse; Surgery; Toxicology; Transplantation; Tropical Medicine; Urology & Nephrology; Virology; Zoology. 253
Slide 8: 254 B.K. Redman and J.F. Merz number of cases alleged or confirmed in the past year; and how they were identified, investigated, and resolved. IRB approval was obtained from the Wayne State University Human Investigation Committee. Results Written policies were located for seven of the fifty journals (Annals of Internal Medicine, Endocrine Review, Gastroenterology, Journal of Clinical Investigation, Journal of the National Cancer Institute, Nature, Physiological Reviews; one is based in the U.K. and six in the U.S.). Five (all U.S.–based) used a definition similar to that of the U.S. Office of Research Integrity (ORI): Research misconduct is fabrication, falsification or plagiarism (FFP) in proposing, conducting or reporting research. (42 CFR, Section 93.103). The remaining two did not provide a definition. Notably, three of the five expanded the definition to include one or more of the following criteria: duplicate publication; misrepresentation of author contributions; failure to disclose conflict of interest; violation of federal, state, and institutional rules; honorary authorship; and animal or clinical research without IRB approval. Most (6 of 7) written policies described the process used to deal with evidence or allegations of research misconduct. The Journal of the National Cancer Institute’s policy is unique in noting that if the author had been sanctioned by ORI, the manuscript would be returned unless a cover letter affirmed that the research was unrelated to the matter for which the author was sanctioned. Three journals used committees or consultants for advice. Nature editors may seek advice about submitted papers not only from technical referees but also on any aspect of an article that raises concerns. Advice may include ethical issues or issues of data or material access. Policies of five journals indicated the allegation should be reported to the author’s institution. A sixth asked the complainant to do so and a seventh indicated the funding agency and the author should be notified. Policies of two journals specified that authors should be notified of the complaint. Only one (Physiological Reviews) explicitly included a right of appeal by the author, to the publication committee of The American Physiological Society (which publishes the journal). Thus, most policies do not provide
Slide 9: Policies on Research Misconduct 255 for investigation of alleged misconduct by the journal, usually because journals have neither the capacity to perform an investigation nor an employment relationship with the author. This finding is congruent with ICMJE and COPE recommendations. Four journals published policies about sanctions, including rejection of the manuscript; publication of errata, retraction, letter of reprimand, requirement that the author publish an apology to correct the record, banning future submissions, dismissal from the society with which the journal is affiliated, and written notification to other professional societies, institutions and funding agencies. Errata, corrections, and retractions may be accomplished by editorial comment drawing attention to the infraction. Although not available on Journal of the American Medical Association’s (JAMA’s) Web site, the American Medical Association’s Manual of Style (1998) provides general information about research misconduct and refers to ICMJE’s policy recommendations and to JAMA’s policy on retraction, which may occur because of research misconduct (Iverson, 1998). In summary, for only seven of the top fifty high impact biomedical journals could publicly available policies be located. When a definition of misconduct is provided, core infractions of falsification, fabrication, or plagiarism are frequently expanded to include duplicate publications, failure to disclose conflicts of interest, and in some cases failure to secure IRB or Institutional Animal Care and Use Committee (IACUC) approval. Nine of fifty editors (18% response rate) responded to an e-mail questionnaire or follow-up phone call interview seeking: verification of existence of a written policy on research misconduct, what precipitated adoption of such a policy or why one was currently being considered; and number of allegations or confirmed research misconduct they had experienced in the past year, how they were identified, and how they were resolved. Two reported having a policy on research misconduct; these responses were congruent with findings of the search for publicly available written policies. While twenty-six of the fifty journals are review journals, only two of the seven journals for which written policies were retrieved were review journals (Fisher’s exact test p>.18). Informal comments and refusals to complete the questionnaire came from editors of these journals with the note that research misconduct is not
Slide 10: 256 B.K. Redman and J.F. Merz applicable to them since they do not publish primary research and the reviews they publish are frequently invited from specific authors. One of the three editors considering adoption of a policy on research misconduct was from a review journal. Seven of the nine editors reported they had a total of 18 (range 1–4, mean 2.6) allegations of research misconduct in manuscripts under review or published by their journal in the past year. Potential cases were identified by readers, other authors, peer reviewers, internal review or contact by ORI. The journals’ investigations included asking the author for an explanation, and if required, referring the matter to the author’s employing institution (2 instances). Retraction of the article was perceived to be a rare event (1 instance reported). Correction was more common (4 instances) because the investigation found an innocent error or as one editor phrased it, “beautification,” rather than misconduct. Two editors without written policies reported that allegations were resolved on a case-by-case basis. Discussion Since publications are the main currency of a scientific career, journals, especially high impact ones, are in a strong position to deter research misconduct. Yet, despite calls for development of journal policies for managing allegations of misconduct, few have adopted them. Although review journals (about half of the sample studied here) would seem less likely to encounter fabrication or falsification, plagiarism could be a problem for all journals. Policies that do exist rarely address all relevant issues: a definition of research misconduct; a process including notification of relevant parties and use of committees/consultants, and right of appeal; and a set of sanctions commensurate with the possible infraction. Some innovative practices by which journals can deal with research misconduct have begun. Editors are starting to use software to identify plagiarism (Taking on the Cheats, 2005). For example, a consortium of otolaryngology, head, and neck journals have agreed to share information about relevant incidents and in egregious cases (after an appropriate deliberative process) may elect to limit the author’s privilege to publish in those journals for a specified period of time. This agreement is expected to limit opportunity for authors who have engaged in unethical practices
Slide 11: Policies on Research Misconduct 257 to simply resubmit their work to another specialty journal where the questionable practices might go undetected (Benninger et al., 2005). One other method that may help deter this practice is requiring authors to submit, along with a manuscript, copies of any previous peer review received on the article. It is reasonable to believe that journals’ adoption of explicit policies is much preferable to taking a case-by-case approach for dealing with allegations of research misconduct. Publicly available policies may have a deterrent effect and can prevent arbitrariness. A decade ago, analysis of handling of a prominent case involving Bernard Fisher precipitated reforms in notification of journals about research misconduct. Problems with some data in the huge multisite trial were not communicated for three years to the New England Journal of Medicine, which published the original article (Anderson, 1994). Journals have an important role to play, but they must ensure that they manage allegations fairly and consistently. Formal adoption of standards, with explicit protection of the rights of whistleblowers and accused scientists, is the best way for journals to proceed. This study has important limitations. Relevant policies may have been missed on Web sites or in journals. The response rate by editors was low. Nonetheless, our data strongly suggests that the leading biomedical journals have not yet adopted policies for managing allegations of misconduct involving manuscripts or published articles. References 42 CFR, Section 93.103(2006). Anderson, C. (1994). How not to publicize a misconduct finding, Science, 263:1679. Benninger, M. S., Jackler, R. K., Johns, M. M. E., Johnson, J. T., Kennedy, D. W., Ruben, R. J., Sataloff, R. T., Smith, R. J. H., Weber, P. C., Weber, R. S., Young, E. D. (2005). Consortium of otolaryngology—head and neck surgery journals to collaborate in maintenance of high ethical standards, Archives of Otolaryngology Head and Neck Surgery, 131:381–382. Budd, J. M., Sievert, M. E., Schultz, T. R., Scoville, C. (1999). Effects of article retraction on citation and practice in medicine, Bulletin of the Medical Library Association, 87:437–443. Iverson, C. (ed.) (1998). American Medical Association Manual of Style, 9th ed. Baltimore: Williams and Wilkins, pp. 103–107. Journals Citation Reports – Science Edition (2004). Philadelphia: Thomson Corp.
Slide 12: 258 B.K. Redman and J.F. Merz Knox, R. (1983). The Harvard fraud case: Where does the problem lie? Journal of the American Medical Association, 249:1797–1807. Locke, R. (1986). Another damned by publications, Nature, 324:401. Lowry, F. (1994). Dr. Roger Poisson: “I have learned my lesson the hard way,” Canadian Medical Association Journal, 151:835–837. Service, R. F. (2002). Scientific misconduct. Bell Labs fires star physicist found guilty of forging data, Science, 298:301. Taking on the Cheats (2005), Nature, 435(7040):258–259. Woolf, P. (1981). Fraud in science: How important, how serious, The Hastings Center Report 11(5):9–14.