Leder: Redelighed i forskning kan ikke tages for givet
Redaktionen
Originalartikel: Sudbøsagen: artikler i førende tidsskrifter på fabrikerede data
Henrik Toft Sørensen
The Sudbø Case
Bibl Læger 2013;205:110-18
In 2005 the Lancet published a Norwegian paper that turned out to be based on fabricated data. In this paper the author describes his experience with scientific fraud when he was asked to review the fast track paper for the Lancet. Surprisingly, the quality of the paper was very poor for which reason the author gave it a low score and recommended rejection. However, a few weeks later Lancet published the paper. Six months later it became known that the first author of the paper, Jon Sudbø, had fabricated the data. The case became a media sensation in Norway and the paper was ultimately retracted. A Commission of Inquiry furthermore deemed Sudbø’s work invalid. He had used fabricated data in several other papers published in leading journals.
Originalartikel: Oversigt over empiriske undersøgelser af hyppighed af videnskabelig uredelighed
Anine Westh Skibsted
Review of empirical studies on the occurrence of scientific dishonesty
Bibl Læger 2013;205:119-34
In this review of the most important empirical studies of scientific dishonesty, the occurrence of dishonesty and proposals for prevention are discussed. Two major cases of research fraud, the American anesthesiologist Scott S. Reuben and the Korean stem cell researcher Woo-Suk Whang, are described. Literature was identified in PubMed, by reference chasing, and on the Danish Committee on Scientific Dishonesty’s (DCSD) webpage. One systematic review described that questionnaire surveys on average find that approximately 2% of researchers reported to have behaved dishonestly, whereas 14% reported to have observed dishonest behavior. Only 0.02% of the papers on PubMed are retracted because of dishonesty. One questionnaire survey indicated that 20% have changed research design, methods or results because of sponsor pressure. Studies show that medical students have trouble identifying dishonesty despite having participated in courses. Both positive and negative correlations between education and dishonesty have been identified. Postgraduate students regard scientific dishonesty as more common than senior researchers who tend to see the problems in a grey area. It is difficult reliably to assess the frequency of scientific dishonesty. On average 2% of the scientists who participated in questionnaire surveys reported to have behaved dishonestly. This number is probably underreported.
Originalartikel: Videnskabens kultur: Hvorfor snyder forskere ikke mere, end de gør?
Klemens Kappel
The Culture of Science: why don't researchers cheat more than they do?
Bibl Læger 2013;205:135 –53.
The paper proposes an analysis of the common idea that severe cases of scientific misconduct call for a cultural change in scientific institutions. The aim of the paper is to characterize aspects of the culture of science and discuss the extent to which they should or could be changed. It is suggested that individual and institutional barriers to scientific fraud or questionable scientific practices are rather few, compared to the incentive that one would expect scientist to have to engage in such practices. So, the main barrier rests in certain aspects of the culture of science, more precisely the internal reward practices found in the culture of science. It is argued that it would be appropriate to strengthen this aspect and that the prime agents in promoting such change would be scientists themselves, not their institutions. Yet, there are certainly more indirect steps that institutions might also consider.
Et billede fra min hverdag
Bente Pakkenberg
Originalartikel: Tobaksindustrien: »Spred tvivl om stærk videnskabelig evidens, og offentligheden vil ikke vide, hvad den skal tro«
Charlotta H. Pisinger
The Tobacco Industry: ‘‘Spread doubt over strong scientific evidence and the public won’t know what to believe’’
Bibl Læger 2013;205:156-81
As part of the 1998 U.S. Master Settlement Agreement, the tobacco industry was forced to disclose millions of previously confidential internal documents which revealed documentation of a close collaboration between the tobacco industry and the academic world. World’s leading doctors and other researchers and highly respected universities have received huge amounts of money, not only to their research projects, but also as consultants and expert witnesses. They have disputed the evidence of harmful health effects of smoking and passive smoking in decades misleading the smokers, the public and the decision makers. The scientific dishonesty includes fabrication of data, falsification, selective publication of studies and effect measures, suppression of independent research and ghost writing. The article describes the research committees founded by the tobacco industry, the different types of projects they sponsored, gives an example of a sponsored study and mentions some important scientists working for them. Also it describes the general strategies used by the tobacco industry to suppress independent tobacco research and promote tobacco friendly research to influence important decisions on e.g. indoor air regulation. Even though many universities now have banned cooperation with the tobacco industry many researchers are still involved in advisory boards and as expert witnesses.
Kvartalets genstand
Morten A. Skydsgaard
Originalartikel: »De sygdomsfremmende virksomheder«
Fokus på fødevareindustrien og videnskabelig uredelighed
Torben Jørgensen
“Industrial epidemics”. Focus on ”Big Food” and scientific misconduct
Bibl Læger 2013;205:184-201
During the latest 40 years the world has faced a dramatic increase in non-communicable diseases, especially obesity and type 2 diabetes. At the same time a major expansion of for-profit corporations producing hazardous products as tobacco, food, sugar, and alcohol has taken place. These products are main drivers for the increase in non-communicable diseases. In this article the historical development of these corporations and the vested interests achieved is described. Numerous studies have provided information of the methods used by corporations to increase sale of their products. The methods are the same for all type of corporations including advertising, lobbying, economic support, law suits, and scientific misconduct. Especially the actions of the tobacco industry are described in details in the literature, but recent studies support that the food industry follows the same track as the tobacco industry. One of the key problems is the very constructs of a corporation. A key protection for a corporation is the limited liability of shareholders and officers, who cannot be held personally responsible for decisions made by the corporation. In this article examples are given on how the food industries manoeuver in the same way as the tobacco industry with huge amounts spent on advertising especially to children, support to politicians and lobbying activities, law suits against reasonable proposal to decrease the intake of unhealthy foods and scientific misconduct with false analyses and interpretation of scientific material. There is an inborn conflict between increasing profit and taking responsibility for public health. If the health of the general population shall be improved it means reduced intake of tobacco, alcohol, food etc., which means less production and reduced income. It seems as if the way forward will be some kind of regulation of corporation from health authorities both national and international.
Originalartikel: Snyd i medicinalindustriens kliniske forsøg
Peter C. Gøtzsche
Fraud in the pharmaceutical industry’s clinical trials
Bibl Læger 2013;202:15.
This article analyses and discusses the role of fraud and misinformation in clinical trials controlled by pharmaceutical companies. The drug industry cheats routinely in their clinical trials. The cheating is highly effective for the marketing of drugs, as it can usually only be detected if independent researchers get access to the trial protocol and the raw data. The difference between an honest and a not-so-honest data analysis can be worth billions of Euros on the world market. To some extent, the medical journals contribute to this unfortunate state of affairs, partly because of their own financial conflicts of interest, partly because the editors are not sufficiently critical towards industry controlled trials. The cheating has dire consequences in terms of waste of taxpayers' money and the loss of lives of thousands of patients. In our current system, we have allowed the drug industry to be its own judges. This is unreasonable and harmful for public health. Testing of drugs should be a public enterprise.