Is research fraud becoming more prevalent? Or are more people just paying attention? Last week, Dr. Scott S. Reuben, an anaesthesiologist at Baystate Medical Center in Massachusetts, admitted that he fabricated the data for his research on post-surgery pain management. Dr. Steve Schafer, editor-in-chief of Anesthesia & Analgesia, was quoted in a New York Times article saying:

“[Reuben] was one of the most prolific investigators in the area of postoperative pain management . . .[His fraud] sets back our knowledge in the field tremendously.”

How was Reuben’s fraud spotted? Questions were raised by his institution after he filed two abstracts and it became apparent that he neglected to obtain ethics approval to conduct studies on human subjects.

I’ll bet that Baystate heaved a sigh of relief to know that he made-up the data rather than created a huge research liability for the institution by doing the study without ethics approvals. (Although, I did a quick Google blog search and already found one personal injury law firm posting information about Reuben on its site.)

Research fraud not only damages the reputation of the individual, it puts their entire body of research in question. Other collateral damages include: taints on the research ethics of colleagues and trainees that participated in the studies, publications or other collaborations; and derailment of the activities of other research teams who based their hypotheses on the findings of the discredited researcher.

Personally, I find the potential damage to trainees the most the most offensive. It creates a new generation of researchers who have been mentored in the methods of fraud, with lowered ethical standards and lacking skills in proper experimental design and analysis.

Interestingly, Reuben is also accused of publishing forgery. Dr. Evan Ekman told Anesthesiology News that his name appeared in two of the retracted papers even though he did not contribute to the manuscripts.

Fraud in research is not particularly new. In 2005, the journal Nature published the results of an anonymous survey of ~3,200 scientists where 0.3% admitted to falsifying research data at some point in their careers and 6% admitted to not presenting data that contradicted their previous research. In an era of rapid fire publication, it would be interesting to try to determine how many papers are published too soon.

Once upon a time, before Harms & Company was even a twinkle in my eye, I analyzed a research project. My quality assurance (QA) and research efficiency project came in after the studies were designed, completed and analyzed by the scientific team. I was supposed to look at all of the information (from the data used to generate figures down to the lab notebooks with the observations) to see if I could come up with the same conclusions as the team. To make a long story short – I couldn’t. There were errors in experimental design, errors in transcription, outliers that weren’t discarded, inliers that were removed and data blinding errors. Even when I used the same data sets (after correcting the transcription errors) I still couldn’t state the same conclusions. Research fraud, poor experimental design or simply a lack of knowledge combined with pressure to produce results? I think a combination of all factors played a role.

For more details on the Scott S. Reuben case, see the Anesthesiology News article, “Fraud Case Rocks Anesthesiology Community”.