Professionalism/Luk Van Parijs and Data Fabrication

Background
According to the Public Health Service (PHS), under the U.S Department of Health and Human Services (HHS), research misconduct is defined as fabrication, falsification, or plagiarism in proposing, performing, or reviewing research, or in reporting research results. The United States Office of Research Integrity (ORI) classifies data fabrication as "making up data or results and recording or reported them," as contrasted with data falsification, which involves manipulating, changing or omitting results such that the research is not accurately represented in the record. A 2009 study on the frequency of scientific misconduct found that 1.97 percent of scientists admit to have fabricated, falsified or modified data or results at least once; however, when surveyed about the behavior of their colleagues, reporting rates increased to 14.12 percent.  How Many Scientists Fabricate and Falsify Research? A Systematic Review and Meta-Analysis of Survey Data. (2009). http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0005738#pone.0005738-Saunders1 In 2015, BioMed Central, a major publisher of scholarly medical and scientific articles, retracted 43 papers after an investigation raised suspicions of fabricated peer reviews. Speaking to the broader problem of peer review and the integrity of published scientific journal articles, the Committee on Publication Ethics (COPE) issued a statement in March of 2015 detailing it had "become aware of systematic, inappropriate attempts to manipulate the peer review processes of several journals across different publishers." Given the explosion in published research in recent decades, data fabrication, as a subset of the larger issue of research misconduct, has great implications for society and the modern day professional.

Luk Van Parijs & Data Fabrication
Born in Belgium in 1970, Luk Van Parijs quickly gained recognition throughout various academic and scientific communities as he pursued undergraduate studies at Cambridge University, graduate studies in the department of pathology at Harvard, earned a PhD in Immunology at Harvard, then received a post-doctoral fellowship at Massachusetts Institute of Technology (MIT) and California Institute of Technology (CalTech) Joining the department of biology at MIT in 2000, Van Parijs was promoted to Associate Professor in July 2004. During his time at MIT, Van Parijs published more than 20 papers in the research field of immune system functions and RNA interference (RNAi) technologies, including a high impact paper in Nature Genetics cited 247 times. In August of 2004, members of Van Parijs laboratory in the Center for Cancer Research reported allegations of misconduct to MIT when they "couldn't find or identify some of the data" used for certain publications, reported Alice P. Gast, vice president for research and associate provost. Internal investigations at MIT were immediately conducted, as information from research materials, interviews with respondents of allegations, and others involved in the research were carried out. In October 2005, at the conclusion of the investigation, Van Parijs was dismissed from MIT.



Several external investigations were subsequently conducted, including an investigation by the U.S Office of Research Integrity, given Van Parijs had received federal funding through research grants. Report findings dating back to Van Parijs' graduate work at Harvard, released on January 23, 2009, detailed findings of data fabrication and falsification in nine different grants supported by the NIH. In total, the ORI investigation found "false data in seven published papers, three submitted papers, ...one submitted book chapter and multiple presentations." The scope of Van Parijs' offenses included instances of falsifying graphical data by duplicating box plots to falsely represent immune cell expression behavior and for "falsely claim[ing] in the text... and [figures of published works] that experiments depicting the functional silencing of genes in stems cells...[and] in dendritic cells by lentiviral-mediated RNAi were performed, when they were not." In June 2011, Van Parijs' pleaded guilty to grant fraud in the U.S District Court in Boston and was sentenced to "six months of home detention with electronic monitoring, plus 400 hours of community service and a payment to MIT of $61,117 — restitution for the already-spent grant money that MIT had to return to the National Institutes of Health." As late as May 2012, Immunity issued a retraction involving Van Parijs' 2003 paper "Autoimmunity as the Consequence of a Spontaneous Mutation in Rasgrp1" although it was cited by the ORI's 2009 findings. Before the 2012 retraction, the publication had received 44 citations.

Guidant Case
In the early 2000s, Guidant Corporation was the second largest manufacturer for implantable defibrillators. Following the nationwide release of Prizm 2 DR in 2000, an implantable defibrillator for the treatment of life-threatening heart attacks, Guidant experienced tremendous market growth. However, a serious problem with Prizm 2DR came to light through a public disclosure by The New York Times in May 2005 with the death of 21 year old Joshua Oukrop, a college student from Minnesota. Oukrop had hypertrophic cardiomyopathy, a genetic disorder affecting the normal function of his heart, and died suddenly after his Guidant defibrillator experienced massive electrical failure during delivery of a shock, a result of deteriorated insulation. Only after public disclosure by The New York Times, Guidant revealed that company officials had known that Prizm 2 DR defibrillators repeatedly short-circuited for more than three years, but had decided not to alert doctors or patients. Senior officials decided it was neither necessary nor advisable to inform physicians or issue an official advisory, alert, or recall. Within management, critical decisions were driven by statistical projections created by low-level product control engineers presenting a false assurance of reliability, devoid of medical oversight, which was then uncritically accepted by company executives. The company continued to sell over 4,000 potentially defective defibrillators, which led to the recall of over 300,000 defibrillators over the next 12 months. A total 7 deaths were reported from defective Guidant defibrillators; however, these rates were largely underestimated because the condition of implanted defibrillators is rarely examined fully after death. In summary, Guidant intentionally kept this problem as non-disclosure. Even though company officials knew that their defibrillators malfunctioned and could result in the death of a patient, they failed to notify Food and Drug Administration (FDA), physicians, or patients.

Space Shuttle Challenger
The devastating case of the space shuttle Challenger disaster can be related to data falsification. On the morning of the launch on January 28, 1986, National Aeronautics and Space Administration (NASA) was debating whether to proceed with the launch of space shuttle Challenger. On the morning of the launch, the temperature was between 26-29 degrees Fahrenheit, which was more than 30 degrees colder than any previous successful launch temperature. NASA evaluated only a subset of the available data for O-ring damage index against temperature, which included only failed launch data. By failing to evaluate the successful launch data, NASA officials decided to proceed with launch given the uncertain risks imposed by extremely low temperatures. The Rogers Commission, a special investigation commission appointed by president Ronald Reagan, found that the O-ring had no test data to support any expectation of a successful launch in such conditions. Also, NASA managers disregarded warnings from engineers about the dangers of launch posed by the low temperature. Overall, this particular case is an example of unintentional data falsification. However, while unintentional, the result was catastrophic as the lives of seven astronauts were lost. Therefore, all data should be presented and examined to avoid misrepresentation of the data.

Implications for the Modern-Day Professional
The potential academic and financial gains in each of these cases are somewhat obvious motives. However, Luk Van Parijs committed the most blatant offense. His results were entirely fabricated, while Guidant's falsification is more relatable to not telling the whole truth, rather than a lie. The challenger case may not have been intentional at all, but data was poorly represented by omissions. The consequences of Van Parijs' actions were small relative to the deaths in the Guidant and Challenger cases; however, by fabricating and falsifying published results, his research had the potential to misguide the work of future researchers and result in costly, ineffective treatments. However, the case of Van Parijs demonstrates the importance of proactive, internal reporting. The Guidant and Challenger deaths may have been prevented if professionals had felt empowered to come forward and high-level officials were willing to consider their input. Those that reported the Van Parijs data fabrication were also invested in his research. The members of his lab sacrificed their jobs and were set back in their academic careers. This highlights the role of a professional to hold others and themselves accountable, despite the costs.

Loyal to Whom?


In a study at the University of Virginia (UVA), 21.7 percent of students reported witnessing an honor code violation in the classroom. 98.3 percent of these students did not report the honor offense. About 56 percent of the 98.3 percent that did not report witnessing an honor offense stated concern for the consequences of the other person as the primary reason. . These students had much less to lose by reporting these offenses than the members of Van Parijs's lab. Those that reported Van Parijs's offenses had potential graduate and doctoral degrees, as well as jobs that revolved around his research. These individuals defied their inherent loyalty to Van Parijs at the expense of their own careers.

In both of these cases, individuals are placing loyalty in others that are being dishonest. Was it disloyal for the members of Van Parijs's lab to confront the problem? Is it possible to be loyal to yourself while being loyal to another when the two are in conflict? John Corvino proposes that "loyalty is only a virtue to the extent that the object of loyalty is good." This view suggests that loyalty is not always good. In this view, one could weigh the virtue of loyalties and act accordingly. A professional should consider the object of their loyalty before trusting blindly.

Behavior Reinforcement
The University of Virginia case study brings up a second major issue: accountability and behavior reinforcement. It has been shown that the likelihood an individual cheats increases the more he or she gets away with cheating. The frequency of unreported cheating at UVA offers a possible explanation of research misconduct. It is possible that Van Parijs and other cases of academic fraud may be the result of a system that protects such behavior. By not reporting the relatively harmless cases, such as copying homework, future cheating behavior is reinforced.

The Skeptical Reader
Peter Norton, a professor at the University of Virginia, often advises students to "write for the skeptical reader" and provide supported, well-founded evidence for all contentions. This statement implies that there are skeptical readers. Writers have the responsibility of making a case, while readers have the responsibility of confirming that the material is in fact true.

Van Parijs lab's members were skeptical enough to discover a lack of evidence. This highlights the responsibility of every individual to be skeptical. It is in everyone's best interest to question authority figures, the media, friends, and colleagues. By questioning Van Parijs, the individuals may have lost their jobs, but they prevented further futile work. Questioning does not have to be confrontational, but separating the truth from lies gives an individual greater autonomy and control over future outcomes.