Articles on Law, Science, and Engineering

News Fraud Fuels Science Fraud

by Charles Walter and Edward P. Richards, III, 13 IEEE Engineering in Medicine and Biology Magazine #2, pg. 279 (Apr/May 1994)

In the September, 1992 issue of EMB Magazine, we outlined the dangers of Hubris and how entrenched interests already vested by the current system of grants, tenure, peer review, and graduate education can play into the hands of zealots and politicians. In the last issue of EMB Magazine, we outlined the emerging "Dingellism" in which zealot scientists team up with politicians to root out evil that they claim threatens the very foundations of the Republic.

At the time that article was written, the Department of Health and Human Services ("HHS") had found Dr. Mikulis Popovic innocent of scientific misconduct and had dropped all charges against Dr. Robert Gallo. While there was much speculation by the press that HHS had changed the standard for scientific misconduct to let the scientists off, the formal rulings of The Research Integrity Adjudications Panel ("Panel") were not available, making it impossible to evaluate these charges. We have now obtained the formal rulings from the Panel. As discussed below, the Panel did not adopt a new standard for scientific misconduct. Instead, it found that the Office of Research Integrity ("ORI") had engaged in what could only be described as a witch hunt.

Man Bites Man

When a Research Integrity Adjudications Panel ("Panel") of the HHS Departmental Appeals Board ("Appeals Board") recommended the 3-year debarment of Paul F. Langlois on May 6, 1993[1], there was little press coverage. Dr. Langlois was a research fellow at the National Institutes of Health ("NIH") who had received financial support from an NIH Intramural Research Training Award. Dr. Langlois' case is the only debarment proposed by NIH that was upheld by HHS. Apparently, however, the actual recommended debarment of a relatively unknown research fellow was of little interest to the press. Yet, Dr. Langlois' case is a classic example of a scientific mentor who became concerned about a research fellow's data, attempted to reproduce them without success, and confronted the individual. The system worked, so it is not news.

The Langelois case may also represent the current limit of self-regulation in scientific research--a research fellow investigated by his employer and debarred after confessing that he did engage in misconduct (The appeal concerned the length of the disbarment, not the merits of the case for disbarment).

Would the same system catch a guilty laboratory chief at NIH or a research star professor at some university? Probably not. But, a new system combining the efforts of congressional subcommittees, zealot scientists and the news media is emerging. It is indeed a blunt instrument for investigating research misconduct[2]--well-suited to catch the big fish, albeit randomly, the guilty and the not guilty.

The New McCarthyism

Political demagoguery requires the cooperation of the press. The absence of articulated standards for professional conduct in research is the crucible in which inaccuracies, half-truths and unverified information from "press releases" reach sufficient temperatures to transmute honorable men and women into victims and dishonorable ones into media stars. Hubris and setting standards that fail to address the full range of legitimate public concerns merely validate the perception that misconduct is rampant, thereby creating more media stars and more victims.

Consider the cases of Dr. Mikulas Popovic and Dr. Robert Gallo. Dr. Popovic was exonerated by his Panel [3]. Dr. Gallo's appeal was never even heard by the Appeals Board because his accusers at NIH withdrew their recommendation for debarment following the findings by the Popovic Panel.

Dr. Popovic was the first listed author and principal researcher for the experimental work reported by him, Dr. Gallo and others in an article in Science entitled "Detection, Isolation, and Continuous Production of Cytopathic Retroviruses (HTLV-III) from Patients with AIDS and Pre-AIDS."[4] This paper and three companion papers published in the same issue of Science are regarded by many as a monumental work of modern science. The paper in question is regarded as a germinal work, perhaps the most important paper of this century in virology. Together, the four papers established that a retrovirus was the cause of AIDS. The Popovic paper described the cell system that permitted more detailed studies of the AIDS virus as well as the development of the blood test needed to ensure a safe supply of donated blood to patients in need. As a result of the misconduct charges and the surrounding publicity, Dr. Popovic has been unable to engage in research in the United States for over three years.[5] Meanwhile, AIDS threatens humanity.

Man Eats Man

Human feeding frenzies can be more violent than those conducted by other predators. In contrast to Dr. Langlois, Dr. Gallo and Dr. Popovic are high-profile scientists. The accusations of misconduct against Dr. Popovic and Dr. Gallo attracted the attention of Congressman Dingell (D. Mich.) and his Subcommittee on Oversight and Investigations. Congressman Dingell's calls for action resulted in the formation of fraud-hunting teams of scientists at NIH, first called the Office of Scientific Integrity ("OSI") and then the Office of Research Integrity ("ORI"). In their quest for fairness from their fellow scientists at NIH, Dr. Popovic and Dr. Gallo attracted the attention of the national news media. According to one of the more balanced articles on the subject ("AIDS TEST CREATOR CLEARED OF CHARGES" byline by Joyce Price in the November 13, 1993 Final Edition of the Washington Times at A1), the controversy involving Dr. Gallo subsided in 1987 for about two years until charges were revived in 1989 following a lengthy investigative report published in the Chicago Tribune.

When Dr. Popovic's Panel vindicated him, some journalists misrepresented the situation by calling it "a clash of cultures" between lawyers (Appeals Board members) and scientists (NIH fraud hunters) (See "SCIENCE AND LAW CLASH OVER FRAUD-CASE APPEALS" Bylined by Phillip J. Hilts in the November 8, 1993 Final Late Edition of the New York Times at 10B), when in fact the HHS lawyers were merely saving accused scientists from other scientists empowered by government and prodded by congressional witchhunting to conduct a search and destroy mission against high-profile victims. Even if they had happened to catch a guilty one, it would not have been worth the price. Exposure of scientific misconduct by political demagoguery and journalistic misconduct is not what the men who crafted the First Amendment had in mind.

The Chicago Tribune continued its coverage of Dr. Gallo in an article headlined WHEN SCIENTISTS, LAWYERS ARGUE, JUSTICE IS THE LOSER and bylined by John Crewdson in the November 14, 1993 Final Edition of the Chicago Tribune at 1C. The Tribune article states,

On Friday the Office of Research Integrity, ... reluctantly withdrew its misconduct finding against Gallo. The ORI declared that the HHS appeals board's verdict in the Popovic case, and the new standard of proof imposed in both cases, made it "extraordinarily difficult for ORI to defend its legal determination of scientific misconduct regarding Dr. Gallo."

Because of the appeals board, the ORI must now prove "deliberate intent to deceive" on the part of a scientist it says is responsible for a false statement, and that the false statement has a "material or significant effect on the research conclusions of the paper." There can also be "no possibility of honest error."

That particular standard, one investigator noted, goes far beyond the "preponderance of the evidence" standard by which most civil cases are judged, and is even more rigorous than "beyond a reasonable doubt," the standard applied in most criminal cases.

The ORI news release quoted in the Tribune article suggests that Dr. Popovic really did make false statements in his Science article, but escaped a finding that he had committed scientific misconduct due to unexpected new definitions and standards imposed by the Appeals Board. By extension, the media frenzy suggests that Dr. Gallo is also getting off on a technicality.

The Tribune article asks, "What is going on here?" Mr. Crewdson's answer is that the federal government is struggling, so far without much success, to deal with misbehavior by scientists.

Unfortunately for Dr. Popovic and Dr. Gallo, the journalists failed to follow up on the information provided by ORI.

It is surprising that an experienced reporter would not investigate ORI's claim that the proof required in a civil proceeding was set at a standard as high or higher than that used in criminal cases. Since there is always a possibility, however remote, of "honest error," common sense dictates that no tribunal would require a showing that there was "no possibility of honest error." If Dr. Popovic's Panel had actually done so, that would be news. And, if an element of scientific misconduct was that the intent to deceive must be "deliberate," then, in addition to being premeditated, the false statement would have to have been considered in a dispassionate manner, pondered upon for a long period of time (especially according to a preconceived design), and its consequences weighed carefully during prolonged meditation. Perhaps "a deliberate intent to deceive" is required for journalistic misconduct, but common sense dictates that no tribunal would require a "deliberate intent to deceive" as an element of scientific misconduct. Any intent to deceive deviates from accepted practices in research.

What is going on here?

Dr. Popovic's Panel found that no law or regulations established a specific definition of scientific misconduct against which to measure Dr. Popovic's conduct. Therefore, his Panel used, as a limit on the scope of its determination of scientific misconduct, the definition of scientific misconduct adopted by the Public Health Service ("PHS") in 1989:

fabrication, falsification, plagiarism, or other practices that seriously deviate from those that are commonly accepted within the scientific community for proposing, conducting or reporting research. It does not include honest error or honest differences in interpretations or judgments of data.[6]

Perhaps the greatest problem facing Dr. Popovic was that ORI scientists had their own idea as to what research practices deviate seriously from those commonly accepted within the scientific community. While his Panel found that ORI failed to prove that Dr. Popovic deviated seriously from commonly accepted practices, Dr. Popovic was clearly at the mercy of the Appeals Board to ascertain the metes and bounds of deviant practices. What was needed, of course, is a set of clearly articulated standards to serve as guidelines for conduct. Another benefit of articulated standards would be that each scientist accused of misconduct does not have to re-establish the scope of acceptable research practices.

If all this doesn't motivate scientists to set up clear standards of conduct and enforce them, nothing will!

Another claim that Mr. Crewdson and other reporters failed to check was ORI's claim that the Appeals Board had changed the burden of proof required for a showing of scientific misconduct. The Appeals Board, including Dr. Popovic's Panel [7], has consistently applied the standard of proof required by Part 50 of Subchapter D of Chapter I of Title 42 of the United States Code, namely "a preponderance of the evidence." Basically, this means that the weight of the evidence supporting misconduct must be slightly greater than the evidence against misconduct. The standard for proving a criminal offense is "beyond a shadow of a doubt."

What actually happened is that the scientists at ORI used their own definition of scientific misconduct--rather than the definition adopted by the PHS in 1989. Throughout the Popovic proceedings and the earlier proceedings dismissing the misconduct charges against Rameshwar Sharma[8], ORI maintained that any untrue statement in a scientific paper or grant application amounts to scientific misconduct. This definition is incompatible with the PHS definition which excludes "honest error." Even after Dr. Popovic's Panel ruled in his favor, according to the Chicago Tribune article, Dr. Lyle Bivens, who Mr. Crewdson describes as a "the ORI chief, a mild-mannered psychologist" was still saying,

"The scientific community," Bivens said, "has a low threshold of tolerance for false statements." ORI and the board, he said, were in a "fundamental disagreement" over "the importance of clarity, accuracy and honesty in science."

Perversely, the Tribune article attributes these words to the same individual who claimed that ORI was dismissing the misconduct charges against Dr. Gallo because the Appeals Board changed the burden of proof from a preponderance of evidence to a new standard that was impossible to meet.

Nor did Dr. Popovic's Panel rule that in order to prove scientific misconduct, a false statement must have a material or significant effect on the research conclusions of the paper and that there must be no possibility of honest error. His Panel did find that, since the data in question were relatively insignificant and had no effect on the research findings reported in Science, there was no persuasive evidence that Dr. Popovic had a motive to falsify such data[9]. If a motive had existed, it would have been relevant in attempting to prove that a false statement resulted from something other than "honest error."

Perhaps most importantly, the Panel found no evidence that Dr. Popovic falsified the data in the article, irrespective of any possible motives for doing so:

One might anticipate that from all this evidence, after all the sound and fury, there would be at least a residue of palpable wrongdoing. That is not the case. On reviewing this lengthy record and all the evidence and argument related to each of ORI's allegations, and notwithstanding the vigorous efforts of ORI counsel, we find that ORI was simply unable to prove by a preponderance of the evidence that Dr. Popovic is guilty of scientific misconduct -- even under standards first promulgated years after the paper was published.[10]

The bottom line? The Panel was quite clear:

Thus, even if we agreed with ORI (which we do not) that unintentional errors in insignificant details in the paper would constitute scientific misconduct under the 1989 definition, we would not reach a different result here.[11]

If Dr. Bivens or Mr. Crewdson were in the dock, it would be appropriate to ask them how they reconcile their acts and omissions with the spirit of Section 50.103(d)(13) of Subchapter D which requires PHS awardee and applicant institutions to undertake diligent efforts to restore the reputations of persons alleged to have engaged in misconduct when the allegations are not confirmed. HHS and the news media should do no less.

Conclusions

Our thesis is that the threat posed to science might still be counteracted if researchers would establish and enforce standards for their own conduct. However, there will be no meaningful standards for the research profession until the profession clearly articulates its goals and how those goals relate to the goals of individuals within the profession.

There is serious division amongst research professionals about the goals of research and how they relate to the goals of individual researchers. This division is found amongst all professionals, including lawyers and physicians, whose activities require unique skills and knowledge and who practice those skills within an economic system where monetary rewards are potentially limitless. And like some lawyers and physicians, some scientists feel that any act that places individual goals ahead of the goals of the profession and is not purely defensive amounts to professional misconduct. On the other hand, greed is a human trait, and scientists, lawyers and physicians are human beings. Some distinguish greed and self-interest with varying points of demarcation, and feel that the goals of their profession may be compromised a bit for their self-interest, but not, perhaps, for their greed. Others don't even bother with such niceties.

However, for research professionals, the priority of individual goals over professional responsibilities has widespread political implications because most research is supported directly by tax dollars in a quasi-socialized funding system that tries unsuccessfully to incorporate competition as its means of quality control. Fortunately, the evolution of research funding is not governed by complex Darwinian principles. If scientific research is for the public good, the public will support it but every participant will be expected to work toward achieving its goal, not to compete. If, on the other hand, the goal of research is a valuable patent, the public may support it but expect ownership of the patent rights. However, if its goal is the enrichment of the researcher, the researcher should pay the bills. Scientists must identify which of these goals apply to scientific research, and then establish standards of conduct for achieving them. The funding mechanisms are determined by the goals and the standards, not vice versa.

ENDNOTES

1. Department of Health and Human Services Departmental Appeals Board Research Integrity Adjudication Panel, In re Dr. Paul F. Langlois, Decision No. 1409, Docket No. A-93-30 (May 6, 1993). [return to text]
2. Healy, Bernadine (1993). 329 N. Engl. J. Med. 725-727.[return to text]
3. Department of Health and Human Services Departmental Appeals Board Research Integrity Adjudication Panel, In re Mikulas Popovic, Decision No. 1446, Docket No. A-93-100 (November 3, 1993) (hereafter Popovic).[return to text]
4. Popovic, M., M.G. Sarngadharan, E. Read and R.C. Gallo (1984). 224 Science 497-500.[return to text]
5. Joyce Price, AIDS test creator cleared of charges, The Washington Times, November 13, 1993, Part A; Pg. A1, Saturday, Final Edition. [I do not want to use the Miskin piece until it is published. I do not have her permission.][return to text]
6. Popovic at 10.[return to text]
7. Popovic at 3.[return to text]
8. Department of Health and Human Services Departmental Appeals Board Research Integrity Adjudication Panel, In re Rameshwar K. Sharma, Decision No. 1431, Docket No. A-93-50 (August 6, 1993).[return to text]
9. Popovic at 32, 60, 68.[return to text]
10. Popovic at 1.[return to text]
11. Popovic at 11.[return to text]