HOME

Table of Contents

Scientific Misconduct: Science and Lies Without the Videotape

    Examples of Scientific Misconduct

    Brief History of Scientific Misconduct

    Tips for Detecting Scientific Misconduct in the Literature

 

Scientific Misconduct: Science and Lies Without the Videotape

The 1993 landmark Supreme Court ruling of Daubert v Merrell Dow Pharmaceuticals, Inc. made the judge the "gatekeeper" for accepting or rejecting scientific testimony. This gatekeeping authority was extended and reaffirmed in the Kumho Tire Co. v Carmichael Supreme Court decision in 1999. As previous editions of MTI Review have made clear, while there is little contention over the authority of the judge in the courtroom, volumes have been written and the debate continues to rage over what it is that makes scientific evidence relevant and reliable to a specific case. However, there is one area where just about everyone would agree should be excluded by the court--when the expert opinion proffered is based on falsified or fraudulent findings. The problem arises in identifying the tainted literature and those guilty of scientific misconduct, particularly when it involves experts employed in a case or the use of science in the courtroom that has relied, perhaps unwittingly, on science based neither on sound principles nor acceptable standards of ethics.

Examples of Scientific Misconduct

The following two examples taken from the Office of Research Integrity records illustrate how easily fraudulent or falsified data can (and all-too-often do) slip into the literature canon and stay there:

A 1993 investigation found that a scientist fabricated and falsified data in research which had been funded by a grant from The National Eye Institute and The National Institutes of Health. Primary data were missing for almost half of the figures and tables in a series of published papers and manuscripts prepared by the scientist in question. Many instances of data fabrication and falsification were found, including: (1) presentation of data for cell counts that never were performed; (2) indication of multiple data points when, in fact, only a single data point was obtained; (3) elimination of the highest or lowest values in sets of experimental readings; (4) alteration or transposition of data to achieve a desired experimental result; and (5) misrepresentation of the time intervals at which data were collected. By the time this was discovered and the investigation completed, the scientist had published the falsified results in four different journals. The journals were notified to retract the articles, but not until several years after publication.
In 1995, an investigation uncovered that a biochemist engaged in scientific misconduct in biomedical research by intentionally falsifying and fabricating data and claims about the purported effects of electric and magnetic fields (EMF) on cells. This scientist published his fabricated results in 1992 in two scientific journals. The published research findings, which were supported by a grant from the National Cancer Institute and The National Institutes of Health, reported data indicating that EMF exert a biological effect by altering the entry of calcium across a cell’s surface membrane. These claims were potentially very important when published because they purported to link EMF and calcium signaling, a fundamental cell process governing many important cellular functions. EMF, which are ubiquitous forms of radiation that arise from diverse sources such as power lines, home wiring, and household appliances, have been a public concern for potential health effects in humans. The journals were notified that the articles should be retracted. [However, by then, much time had elapsed leaving many unsuspecting scientists attempting to build upon the falsified data for future studies.]

One would naturally surmise that since the parties were found guilty the falsified results would have been retracted and removed from the scientific literature by now. However, this is not the case as both authors still have articles indexed in numerous databases on the Internet and in PubMed, the huge database holdings of the National Library of Medicine which so many scientists rely on in performing their research. In fact, as of this writing, the scientist in the second example is featured as the author of 29 articles in PubMed--most of which reference back to the fabricated results of his earlier work. Consequently, lingering unsound science is not just a problem for the legal community; it is becoming a constant concern for everyone as unethical conduct permeates the sciences. It takes its toll on the majority of researchers that conduct their work in a responsible manner causing them to waste time, effort, and limited financial resources on unsound studies and ultimately results in an erosion of public confidence in all scientific findings. It also costs the federal government and American taxpayers, the single largest provider of research funds for scientific research, millions if not billions of dollars a year, and most frighteningly of all, it jeopardizes the very health and well being of everyone. Scientific misconduct is fraught with serious consequences indeed.

Brief History of Scientific Misconduct

When did we arrive at such a position in the sciences? In the early 1980's both federal funding sources and the scientific community became acutely aware that actions of misconduct in the previous twenty years had caused a watershed change in scientific affairs. Both government officials and scientists realized that more oversight was needed as reports of plagiarism, fraud and falsified data intensified. Scientific societies strengthened their standards and Congress passed the Health Research Extension Act in 1985 which required that all institutions awarded federal research funds establish "an administrative process to review reports of scientific fraud and report to the Secretary any investigations of alleged scientific fraud which appears substantial." From this came the present Office of Research Integrity (ORI) whose mission is to oversee government funded research and ferret out instances of scientific misconduct and mete out suitable punishments. The last decade has been a particularly busy one for the ORI. Between 1992 and 2000 there were 223 investigations of possible scientific misconduct which uncovered 111 confirmed cases that mainly involved fabrication and falsification of data. To aid identification of fraudulent work the ORI publishes some of the names of those found guilty over the last five years and brief information detailing their fabrication and fraud on its website.  A Freedom of Information Request (FOI) is required to obtain a complete list of names and details of the cases.

Identification of Scientific Misconduct

Although obtaining the list is an important step in keeping tabs on scientific misconduct, one needs to remember that the ORI covers only federally-funded research in the US, not worldwide research or privately funded studies. As incomplete as it is, the list should always be at a researcher’s elbow when doing database research. The larger problem of identification of fabrication and fraud would be made somewhat easier if the fraudulent published work was labeled as such in the databases such as PubMed. The National Library of Medicine has the largest collection of medical or scientific records in the world and NLM’s policy is to retract only those articles that have been formally retracted by their author, academic or institutional sponsor, editor or publisher. Therefore, even if a researcher has published an article with falsified data, the published work remains as a part of the scientific canon if the work has not been formally retracted by the researcher responsible. In the main, this results in a "let the buyer beware" reality when it comes to assessing the validity and reliability of the science catalogued in the database. PubMed is not alone in its policy as most of the scientific databases are similarly noncommittal and reactive rather than proactive in their policies toward scientific misconduct. 

Tips For Detecting Scientific Misconduct In The Literature

As scientific professionals, at MTI we frequently encounter the problem this presents and to help our readers sort the sound science literature from the unsound we suggest the following:

1. Always take note of the "errata" that is signified in databases such as PubMed and investigate to see what has been corrected. "Errata" means that a mistake of some sort was made in the initial publication and signals a change or emendation has been made to the earlier article. Traditionally, errata referred only to typographical or printing errors but recently it has come to include factual errors and even instances of fraud.
2. Some citations also have "comments" appended to the citation. This can sometimes represent a red flag and should be followed-up. Though sometimes the "comment" is an endorsement or healthy scientific skepticism of the scientific finding of the research, often the "comment" will make reference to the validity of the methodology, the results, etc. Since even the federally funded research oversight provided by the ORI sometimes takes several years, a careful scrutiny of the immediate "investigation" taking place by fellow scientists of one of their colleagues can sometimes give you a "heads up" that the research findings need to be approached with caution.
3. At a very minimum obtain the complete list of names of those ORI found guilty of scientific misconduct and periodically check the ORI website for updates. (File a FOI request with ORI or       MTI will provide the complete list of 1992-June 2001 at no charge on request.)
4. As indicated in our suggestions with "errata" and "comments," be sure that you have a thorough search performed of the researcher and the findings, including online and manual tracking of bibliographical references.
5. Obtain an expert service such as MTI to make sure you are basing your case on sound science that will both satisfy the strictures of Daubert and keep the opposition honest.
 

Conclusion

As the pace of science continues to hurtle at breakneck speed it is becoming ever more difficult to isolate the perpetrators of fraud. In the old world of scientific research, science most often moved incrementally, reviewing and validating as it moved slowly forward to solid findings and only then would be published in a public forum. Scientific practice today, on the other hand, frequently places such an importance on publishing that the science itself is often eclipsed. As tenuous as the situation with scientific misconduct has become, the problem has been exponentially magnified by the instantaneous capabilities of publishing totally unreviewed findings on the Internet. This frenetic desire for publication along with the often cutthroat competition that exists for research funding, coupled with the overall erosion of scientific ethics, has opened a plethora of Pandora boxes. The "gate" that the Daubert and Kumho decisions directed judges to "keep" also threatens to open beyond a court’s watchful eye. For regardless of how vigilant an individual judge might be in exercising restraint over a question of scientific reliability and relevance, being unaware of scientific misconduct that is relevant to a particular case can only serve to make "gatekeeping" itself an inexact science. Only by adhering to careful attention to detail and obtaining qualified expert assistance that researches not only the science but the scientists who are responsible for it can you ensure that your case has the sound science that will both pass court mandated strictures and possess unassailable standards of scientific ethics.

Complete bibliography available on request

 

©  2001 MEDICAL & TOXICOLOGICAL INFORMATION

MTI REVIEW is a publication of Medical & Toxicological Information, A Division of NICOM, Inc. providing scientific information to non-scientific communities.  Entire contents copyright by MTI.  No other publication or reproduction rights granted without the express written permission of the publisher.

Copyright ©1996- 2005  Medical & Toxicological Information (MTI) A Division of NICOM, Inc.   All rights reserved
703 684-4636       703 MTI-INFO

Home          About MTI             Contact MTI             News/Pub            Services