In the recent BSI careers survey, 5% of respondents said that their scientific careers had been significantly impacted by the scientific misconduct of others. This figure is alarming enough but when you pause to consider how many types of fraud there are, the scale of the issue could be greater.
Fraud or scientific misconduct includes fabrication of data, falsification of data (including data selection and image manipulation), plagiarism (including self-plagiarism and use of other people’s data/ ideas), failure to meet ethical obligations such as obtaining patient consent, misuse of research funds, misrepresentation of data by, for example, not disclosing relevant findings, making inappropriate claims to authorship or failing to include an author who has made a significant contribution. For a more detailed list, there are excellent resources on websites at the University of Leicester1 and Enago.2
Societal implications of scientific fraud
Not only can scientific misconduct impact on researchers, but it can have major societal implications. For example, in 1998, the now discredited gastroenterologist Andrew Wakefield published a fraudulent paper that claimed that the MMR vaccine was linked to an increase in autism. The impact of these false claims has been widespread, with reduced uptake of the vaccine in the UK and around the world. Indeed, we are still feeling the effects of the resulting vaccine hesitancy with the current measles outbreak in England. The ongoing confusion that this fraudulent paper caused is evident in sites such as ‘Mumsnet’ where discussions boards reveal many concerns as to what the best vaccination strategy is.
More recently, investigations by Ohio State University into Ching-Shih Chen’s work on anti-cancer drugs revealed evidence of scientific misconduct in a large number of his papers, resulting in his resignation from the university and the cessation of a clinical trial using one of the anticancer agents he worked on. The career of Paolo Macchiarini, once heralded as a pioneer of regenerative medicine for carrying out trachea transplants generated by scaffolds seeded with patient stem cells, disintegrated after he was found guilty of scientific misconduct and a documentary that showed seven of the eight patients had subsequently died.
Effect on others’ careers
Careers can also be devastated by the fraudulent activity of others. Alirio Melendez, once a Professor of Immunopharmacology at the University of Liverpool, and previously employed by the University of Glasgow and the National University of Singapore, is thought to have committed multiple cases of fraud across institutions. Such sequential scientific misconduct has both direct and indirect impacts on the careers of other scientists.
Those researchers unable to reproduce fraudulent data may feel incompetent, and their ability to obtain grant and fellowship funding might be impacted. If researchers have the misfortune to co-author papers with a fraudster, they may find that, even though their own scientific integrity was not doubted, they are tarred with the same brush and it could affect their future job prospects. Papers of individuals who collaborated with the fraudster prior to the actual scientific misconduct can also see a drop in their citations as people avoid citing a fraudster.3 More worryingly, some young scientists may choose to give up science altogether.
An increasing phenomenon?
Is scientific fraud on the increase driven by the ‘publish or perish’ ethos in modern research institutions? Could increased scrutiny be responsible for a perceived increase in scientific misconduct? Recent work by Fanelli has addressed some of these questions.4 The good news is that, as no change has been found in the yearly report of scientific misconduct cases to the US Office of Research Integrity, it therefore seems unlikely that scientific fraud is on the increase. While it is true that retractions have increased in absolute numbers, it is likely that this is due to improvements in journal retraction policies. Fanelli has also found no link between paper retraction and gender.
Why might Fanelli’s findings differ so much from the perceived view? One reason is that many of us have focused on high profile cases of individuals, such as Luk Van Parijs and Alirio Melendez in the immunology field, who have repeatedly committed research misconduct and published fraudulent results. Such cases get a lot of coverage, making the issue seem much greater than it is.
Given that scientific fraud occurs, what is the best way to prevent it or ameliorate its impact? Teaching ethics and best practice at undergraduate and postgraduate level would be one approach. A more forensic approach to reviewing experimental data at regular lab meetings would be an additional approach. However, it has to be recognised that the range of techniques and the need to process big data for many publications mean that many authors will not be expert in all areas and will need to depend on the scientific expertise and integrity of their co-authors. It would be interesting to start a conversation about how BSI members think this can be dealt with and how the impact of fraudulent science on research progress and the scientific careers of the innocent can be limited.
BSI Chair of Forum
Emeritus Professor, University of Cambridge
1. University of Leicester Research Code of Conduct 6.2 Types of misconduct in research http://bit.ly/2B6n6AO
2. Enago Academy 2018 10 types of scientific misconduct http://bit.ly/2P9BlYO
3. Hussinger & Pellens 2017 ZEW Discussion Papers, No. 17–051, Zentrum für Europäische Wirtschaftsforschung (ZEW), Mannheim
4. Fanelli 2018 PNAS 115 2628–2631