Thursday, March 27, 2008

Policing of Scientific Fraud

The term "scientific fraud" is used to describe intentional misrepresentation of the methods, procedures, or results of scientific research. Behavior characterized as scientific fraud includes fabrication, falsification, or plagiarism in proposing, performing, or reviewing scientific research, or in reporting research results. Scientific fraud is unethical and often illegal. When discovered and proven, fraud can end the scientific careers of researchers who engage in it. Nonetheless, the substantial financial and reputational rewards that can accrue to scientists who produce novel and important re-search or who obtain certain desired results have induced some scientists to engage in scientific fraud.

Policing of Scientific Fraud

Before 1980, only a handful of accusations of scientific fraud were ever proven. In 1981, however, following press reports of a "crime wave" of scientific fraud, the U.S. House of Representatives conducted the first-ever congressional hearings on the subject. These hearings revealed a wide gap in perception of the magnitude of the problem. Prominent scientists testified that fraud in science was rare; that individual allegations were best investigated on an ad hoc, case-by-case basis; and that government intrusion into the evaluation of scientific fraud would place bureaucrats in charge of declaring scientific truth. Prominent journalists and other critics, in contrast, testified that many cases of scientific fraud had likely gone undetected; that the scientific system of self-policing responded inadequately to fraud; and that the government's substantial financial investment in basic scientific research necessitated undertaking measures to ensure the integrity of the scientific enterprise.

Congress, siding with the critics, enacted the Health Research Extension Act of 1985. The Act required federally supported research institutions to develop internal procedures for handling allegations of scientific fraud, and also mandated the establishment of a new government agency to receive and respond to such allegations. That agency, the Office of Scientific Integrity, was established in 1989, and was renamed the Office of Research Integrity (ORI) in 1992. In 1993, the ORI became part of the U.S. Department of Health and Human Services (HHS). In 1999, the ORI ceased conducting its own fact-finding operations, instead ceding that role to the inspector general of HHS. ORI continues, however, to oversee all scientific fraud investigations. From 1994 to 2000, ORI processed 1,205 allegations of scientific misconduct, and sustained findings of scientific misconduct and/or took administrative action in ninety-five cases.

Fabrication of Data or Physical Evidence

There are several varieties of scientific fraud. Perhaps the most egregious incidents involve "fabrication" or "forgery," for example, situations in which researchers deliberately invent or falsify data, or report results of experiments that were never conducted.

A modern incident of "fabrication" arose in 1981, when Dr. John R. Darsee of Harvard Medical School reported the results of experiments in which dogs with induced myocardial infarction were said to have been injected with experimental heart disease medications. In 1982, however, two investigating committees determined that the reported experiments had never taken place. In a similar case, from 1973 to 1977, Dr. John Long of the Massachusetts General Hospital published several papers in which he claimed to have "subcultured" certain permanent lines of malignant tumor cells taken from patients with Hodgkins disease. In 1980–1981, Dr. Long confessed that these claims were fabricated, and that no such cell subcultures had ever been created.

Occasionally, scientists have gone beyond fabrication of data, and have actually fabricated physical evidence to bolster their fraudulent claims. One infamous example of such possible fakery was perpetrated between 1912 and 1915. Amateur archaeologist Charles Dawson discovered two skulls said to belong to primitive hominoid ancestors of man in the Piltdown quarry in Sussex, England. For decades, the "Piltdown man" was widely accepted by the international scientific community as the "missing link" between human and ape, with the noble brow of Homo sapiens and a primitive jaw. In the 1930s, however, the discoveries in Asia and Africa of other, incompatible hominid fossils cast doubt on the authenticity of the "Piltdown man" skulls. Finally, in 1953, an international congress of paleontologists, relying in part on modern dating technology, pronounced "Piltdown man" a hoax. To this day, however, the identity of the perpetrator(s), whether Dawson or others, remains disputed.

A more recent case involving fabrication of physical evidence involved medical research into skin grafting, a process that can enhance the safety of organ transplantation. During the period from 1967 to 1974, Dr. William A. Summerlin of New York's prestigious Sloan Kettering Institute reported that he had successfully transplanted skin from black mice to genetically distinct white ones. In 1974, however, under pressure from his colleagues, Dr. Summerlin confessed that he had used a black felt-tip pen to darken an area of graft that he had actually transplanted from one white mouse to another. Subsequently, a committee of Sloan Kettering researchers determined that Summerlin had also misrepresented the results of his earlier work on corneal transplantation from human cadavers to rabbits.

Misrepresentation of Experimental Results

More prevalent and more vexing than outright fabrication is the "fudging" or "massaging" of data, in which collected data or mathematical computations are manipulated modestly, so as to make the data appear to conform more closely with the researcher's conclusions. A related offense occurs when researchers "cook" or "finagle" data by reporting only part of their findings, while omitting to report data or experimental results that do not support their conclusions. In one famous example, data may have been "cooked" by Robert A. Millikan, the University of Chicago physicist who was awarded the Nobel Prize in 1923 for computing the charge of subatomic particles called electrons. Millikan's computations were based on data that he obtained by charting the movements of oil droplets across electrically charged brass plates. In the 1913 paper that led to his Nobel Prize, Millikan stated that his data "represented all the drops experimented upon." In 1978 and 1981, however, researchers reviewing Millikan's original unpublished notebooks discovered that Millikan had, in fact, failed to report data from 49 of the 140 oil drop observations that he performed, and that the apparent clarity and elegance of his computations were enhanced by the omissions.

By today's standards, omission of data that inexplicably conflicts with other data or with a scientist's proposed interpretation is considered scientific fraud. In Millikan's time, however, scientific measurement technologies were relatively crude, and scientists commonly exercised judgment concerning which of their observations were valid, and which were tainted by laboratory error. In this regard, Millikan's judgment appears in retrospect to have been sound; his exclusion of the forty-nine "bad" data points did indeed enhance the accuracy of his computations of an electron's charge. Millikan's case thus illustrates the inherent difficulty of drawing a line between scientific fraud on the one hand, and the exercise of creative judgment and the force of conviction that remain integral to scientific achievement on the other hand.

The "baltimore Case"

The difficulty in drawing a line between scientific fraud and honest error was illustrated by a prominent controversy involving another Nobel Laureate, the virologist David Baltimore. In 1975, at age thirty-seven, Baltimore was awarded the Nobel Prize in Medicine for his discovery of retroviruses and their means of reproduction. A decade later, while affiliated with the Massachusetts Institute of Technology, Baltimore coauthored a paper that concluded that a mouse's immune system could be altered by injection of a special mouse gene—a finding that raised the possibility of genetic modification of the human immune system. The data analyzed in this paper were derived from laboratory experiments performed on mice by Dr. Thereza Imanishi-Kari, Baltimore's assistant.

In 1985, Dr. Margot O'Toole, a postdoctoral fellow working in Baltimore's laboratory, was assigned to perform additional experiments on mouse genes for a proposed follow-up paper. When O'Toole could not reproduce Imanishi-Kari's original results, however, O'Toole became convinced that Imanishi-Kari's results were "fudged." O'Toole's suspicions seemed to be confirmed when inspection of laboratory notebooks revealed certain discrepancies with the published results. After Baltimore and Imanishi-Kari stood by Imanishi-Kari's original results, O'Toole complained to an acquaintance, who in turn alerted authorities at the National Institutes of Health (NIH) that Imanishi-Kari might have committed scientific fraud.

O'Toole's allegations sparked further congressional hearings in 1989 and 1990, during which Baltimore vigorously defended Imanishi-Kari's integrity, and argued that the government investigation represented a threat to scientific inquiry. Although no one ever accused Baltimore himself of fudging any experimental data, many scientists initially credited the allegations against Imanishi-Kari, who was declared guilty of scientific fraud by both the House Subcommittee on Oversight and Investigations and the Office of Scientific Integrity of the NIH (now the ORI). Tarred with complicity in an alleged cover-up, Baltimore was forced in 1991 to resign from his recently acquired position as president of Rockefeller University in New York City.

In 1996, however, an NIH appeals board wrote a 183-page opinion analyzing the Baltimore case in great detail. The board fully exonerated Imanishi-Kari of engaging in any misconduct, but noted that the research paper at issue had contained errors that Baltimore and Imanishi-Kari had subsequently acknowledged. With some exceptions, scientists generally were persuaded by the board's disposition, and came to view the government investigations of 1989–1991 as "scientific McCarthyism," spearheaded by reckless politicians who sought to obtain publicity as crusaders by humiliating leading members of the academic establishment. In 1997, Imanishi-Kari received a tenured faculty position at Tufts University, while Baltimore was appointed president of the California Institute of Technology.

Bibliography

Broad, William, and Nicholas Wade. Betrayers of the Truth. New York: Simon and Schuster, 1982.

Kevles, Daniel J. The Baltimore Case. New York: Norton, 1998.

Kohn, Alexander. False Prophets: Fraud and Error in Science and Medicine. New York: Basil Blackwell, 1986.

Steinberg, Nisan A. "Regulation of Scientific Misconduct in Federally Funded Research." Southern California Interdisciplinary Law Journal 10, no. 1 (2000): 39–105.

United States Department of Health and Human Services, Office of Research Integrity. Home page at http://ori.dhhs.gov.

United States Office of Science and Technology Policy. "Federal Policy on Research Misconduct." Federal Register 65, no. 235 (Dec. 6, 2000): 76260–76264.

No comments: