Methods

Top Methods Results Discussion Author & Article Info References

We retrospectively reviewed all applications for gastroenterology fellowships that were submitted in a recent year to the Division of Gastroenterology and Hepatology at the University of Pittsburgh Medical Center. Applications were reviewed with respect to 1) research experience during medical school, residency, or interim time periods; 2) presentations [regional, national, or international]; and 3) publications [abstracts, articles, or letters]. Research experience was confirmed by mention of research in a letter of recommendation from mentors or collaborators, verifiable publications, or verifiable presentations. Presentations were verified by mention of the presentation in letters of recommendation or indication of presentation in the program of a meeting. Publications were verified by the following: 1) a copy of the publication enclosed with the fellowship application; 2) citation in the MEDLINE database; 3) a letter of recommendation from a coauthor; or 4) in the absence of these three items, review of the actual journal (if the journal was not available, a copy of the article was requested through the National Library of Medicine). To allow time for publication of articles listed as "in press," we repeatedly searched for such citations in MEDLINE for at least 1.5 years. We used the following in the MEDLINE searches: the applicant's last name; first name, if there could be confusion about which name was the last; coauthors' last names; title of article; and journal name. When a publication could not be confirmed by any of the above criteria, we determined the existence of the cited journal using Ulrich's International Periodicals Directory [2]. This directory contains information on 140 000 serials published throughout the world and is considered the definitive reference on the existence of journals. We defined misrepresentation as the citation of an article or journal that did not exist or the listing of an article as "in press" that was not subsequently published.

To ascertain whether misrepresented publications were unique to any subgroup of applicants, we also reviewed applications in regard to 1) the applicant's sex, 2) graduation from a U.S. medical school versus an international medical school; 3) country where the applicant received previous medical education; 4) U.S. city where the applicant received medical residency training; and 5) the applicant's grade. The applicant's grade had been previously determined by the Division of Gastroenterology Fellowship Application Review Committee during the initial review of applications. At that time applicants were assigned a grade (A+, A, A -, B+, B, B -, C+, C, C -, D, or F) on the basis of the applicant's letters of recommendation, academic record, residency training program, research experience, life experience, and professional goals. The final grade represented the average of grades independently assigned by two faculty members of the Application Review Committee.

Statistical analysis of data was done using the chi-square test.

Results

Top Methods Results Discussion Author & Article Info References

Demographics

Of the 236 gastroenterology fellowship applications, 208 (88.1%) were from men and 28 (11.9%) were from women. The distribution of applicants by grade is shown in Table 1.

Research

One hundred thirty-eight applicants (58.5%) reported research activities. Research was reported in all applicant grade groups and was confirmed for 91 of the 138 applicants (65.9%) claiming research experience. Because it was not possible to verify research experiences during foreign training unless such research resulted in publications, claims of research during foreign training that did not lead to publications were not included in the analysis of data.

Presentations

Seventeen applicants (7.2%) reported presentations at meetings during their U.S. training. Presentations could not be confirmed for five applicants (29.4%). Because we could not verify presentations at foreign meetings before U.S. training, we did not include such claims in the data analysis.

Publications

Eighty-four of the 236 applicants (35.6%) reported published abstracts or articles or both. Of these 84 applicants, 40 (47.6%) reported abstracts. Published abstracts could not be confirmed for 14 of the 40 applicants (35.0%). Some abstracts might not have been located because programs of meetings that contain abstracts may not have been incorporated into journals or deposited in libraries.

Unlike that of abstracts, the existence of published articles is readily verifiable. Fifty-three applicants (22.4%) reported articles, with the number of articles ranging from 1 to 24 per applicant. Articles were misrepresented by 16 of these applicants (30.2%). Table 1 also shows the distribution of published and misrepresented articles by applicant grade.

There were no statistically significant differences between men (15 of 46 [32.6%]) and women (1 of 7 [14.3%]) or between U.S. medical graduates (2 of 14 [14.3%]) and international medical graduates (14 of 39 [35.9%]) with regard to misrepresented articles. Among graduates of international medical schools, there was no preponderance of applicants with misrepresented publications from any single country. Finally, applicants with misrepresented articles received their U.S. residency training in many cities.

Discussion

Top Methods Results Discussion Author & Article Info References

Review of the credentials of gastroenterology fellowship applicants shows that there are many hardworking and productive applicants. However, the accomplishments of some applicants are inflated or nonexistent.

The definition of fraud as it pertains to science and medicine has been widely addressed [3-5]. Engler and colleagues [4] defined fraud as "involving intentional misrepresentation." In fraud, "the perpetrator makes statements that are intended to induce others to believe things which he or she knows to be false" [4]. The National Academy of Sciences' Panel on Scientific Responsibility and the Conduct of Science has defined criteria for the integrity of the research process [3]. The Panel has divided ethical problems in academia into three categories. The first category, "misconduct in science," includes fabrication, falsification, or plagiarism in proposing, performing, or reporting research. The second category, "questionable research practices," consists of actions that violate traditional values of the research enterprise. The third category, "other misconduct," includes "abuses for personal financial gain or professional advantage" [3]. The National Academy of Sciences notes that the fabrication of bibliographic material is included in the panel's definition of "misconduct in science" (category 1). The misconduct perpetrated by fellowship applicants may also have characteristics of category 3—"obtaining professional advantage."

During the initial review of the gastroenterology fellowship applications, only 2 of the 16 cases of misrepresentation were detected. These were recognized by flaws in the misrepresentation. In the first case Table 2, case 4], the applicant cited several papers on diverse subjects, all published during the applicant's first and second years of postgraduate training. The unusual diversity and productivity during this training period brought attention to this application. It was then noticed that one citation included a journal volume number that did not exist. Further investigation revealed that three of the applicant's four "publications" were nonexistent articles cited in existing journals. In the second case Table 2, case 16), the applicant raised immediate suspicion by reporting the publication of a paper about a basic science topic in a purely clinical journal. Investigation showed that the article did not exist. Neither of these applicants was accepted into the fellowship training program in question.

In our study, we found several types of misrepresentation of published articles among the 16 cases (Table 2). These included listing articles in journals that were found to be nonexistent, listing titles of articles that did not exist in the actual journals, or listing manuscripts as "in press" without citing the name of the accepting journal. Some applicants had no publications; their bibliographies were completely misrepresented. Other applicants had some actual publications but enhanced their credentials by adding nonexistent publications.

Following are several possible hypotheses that may explain why a fellowship applicant would misrepresent information. These are not mutually exclusive.

1. Competitive advantage. An enhanced curriculum vitae may be seen by applicants as a way to increase the likelihood of obtaining a fellowship. However, this does not fully explain the motivation of applicants who were at the top of the ranks or who already had established publication records.

2. Low risk–benefit ratio. The probability of detecting misrepresented accomplishments is small. During the initial review of the 236 applications, only 2 of the 16 instances of misrepresentation of publications were immediately appreciated. Only with close retrospective scrutiny were the other 14 instances noted. Even if misrepresentation is detected, the applicant can still practice internal medicine or seek training in another subspecialty. The benefit of misrepresentation is that an applicant may gain entry into a career that he or she views as interesting or lucrative.

3. Materiality. Acceptance into a gastroenterology fellowship program essentially guarantees completion of the fellowship and entry into a subspecialty that has a higher median income than other medical subspecialties. In a recent year, for example, the median income of gastroenterologists was $205 060, whereas the median income for infectious disease specialists was $143 710 [6].To test the hypothesis that materiality may influence misrepresentation, we reviewed all 23 applications received for infectious disease fellowships at the University of Pittsburgh Medical Center during the same study period for confirmation of research, presentations, and publications. Data on the gastroenterology and infectious disease applicants are shown in Table 3. The two groups are similar with regard to the percentage of applicants who misrepresented accomplishments. Although more misrepresented presentations and articles were found among the gastroenterology applicants than among the infectious disease applicants, this difference was not statistically significant.

4. Universality. The perception may exist that it is the norm to inflate one's achievements on applications and that this does not constitute an ethical problem.

5. Innocent error. An applicant might believe that the submission of a manuscript to a journal qualifies the article to bear the appellation "in press." An applicant may also have incorrectly translated the title of an article published in a foreign journal or the journal name, thus preventing the article from being located in the MEDLINE database or Ulrich's Dictionary [2].

6. Visa status. An applicant who is a foreign national may have a visa requiring the applicant to continue training to remain in the United States. We did not have sufficient information on visa status to draw any conclusions about the relation between this hypothesis and misrepresentation of accomplishments.

7. Mental aberration. Misrepresentation may reflect an underlying psychiatric problem.

Whatever the motivations for applicants to misrepresent their credentials, institutions receiving such credentials should be aware that fraudulence occurs. Other instances of fraudulent curricula vitae have been reported [4, 7]. Two of these cases have received much notoriety. Broad and Wade [7] chronicled the case of Elias A.K. Alsabti. Dr. Alsabti advanced through the academic training system by having a curriculum vitae that contained 60 publications published over a few years. Dr. Alsabti had not conducted much of this research but had cleverly plagiarized existing scientific articles and resubmitted the entire work to another journal under a new title, with himself as an author. A second example was reported by Engler and colleagues [4]. They described the case of Robert Slutsky, whose bibliography included 137 articles published in 7 years of residency training. Forty-eight of these articles were judged questionable, and 12 were deemed fraudulent. Although the misrepresentations detected in our study were not of the scale of the Alsabti or Slutsky cases, it is of concern that misrepresentation is occurring on any scale.

In our study, approximately 30% of applicants reporting academic accomplishments misrepresented these accomplishments. Because of our study methods, this percentage is at the highest end of possibility. However, even if the actual percentage of misrepresentation were one half of that found, the problem would still be a significant one. The findings from two fellowship programs in a single institution cannot, of course, necessarily be generalized to all other training programs in the United States. The programs we studied may have been perceived by applicants as valuing academic accomplishments more highly than other programs. Thus, applicants may have felt increased pressure to exaggerate their accomplishments in their applications to these programs.

Our study raised the vexing dilemma of the duty of the discoverers of misrepresentation versus the issue of confidentiality of information provided on applications. If these discoverers are to reveal their findings, who then should be notified? Should it be the programs where the applicants received their training, those persons who wrote letters of recommendation for the applicants, or certifying and licensing organizations? No guidelines for this situation exist, but they must be developed. Guidelines should define the obligation of those who discover misrepresentation. Guidelines should also define the responsibilities of faculty for preventing misrepresentation, educate trainees about why such behavior is destructive and unacceptable, and facilitate the development of institutional policies on misrepresentation. Finally, guidelines should encourage the development of professional organization policies to address eligibility for certification, licensure, and staff privileges. An important first step in this direction has been taken by the American Board of Internal Medicine in its recent publication Project Professionalism [8]. This project identifies several issues of professional ethics, including misrepresentation. We believe that our study provides a useful algorithm for future studies of misconduct.

In conclusion, research activity or publications, or both, were commonly misrepresented on applications for a gastroenterology fellowship. Misrepresentation was not limited to gastroenterology applicants: Review of the limited number of applications for infectious disease fellowships also showed this problem. We urge the academic community to begin to address and correct the serious situation of misrepresentation by trainees. We recommend that the following steps be implemented: 1) that programs require that copies of all publications and of letters of acceptance for manuscripts in press be submitted with applications; 2) that application forms contain a statement, to be signed by the applicant, that the information provided by the applicant is accurate and that falsification of information will be grounds for dismissal if that applicant is accepted into that training program; 3) that persons writing letters of recommendation verify information such as research activity and presentations; 4) that medical students and residents be taught that misrepresentation of research activity or publications constitutes misconduct; and 5) that institutional and professional organization policies be developed to deal with misrepresentation on applications.