Pressuring junior colleagues to commit fraud

Just say no

What should you do when your boss puts pressure on you to commit fraud or to bend the research evidence toward a more desired outcome? Just say no.

A colleague who wishes to remain anonymous to avoid being persecuted by her former boss told me this story.

During my first meeting with a high-ranked doctor, he told me without mincing words about the modus operandi of his lab: Publications come first and science second because publications bring in grants.

Being a naive and newly baked postdoc, I had no idea what he meant until he asked me to falsify the data. He wanted to know whether a statin drug from Merck could be effective for aortic aneurysm. We used a well-known animal model of aneurysm induction. The mice were fed a high fat diet, and one group received a statin, the other a placebo.

At the end of the experiment, there were more dead mice in the statin group. My boss told me to ignore the dead mice because it would not look good for the drug.

I consulted with the most senior person in the lab who was moving on to a different position and she understood why I was upset. I ended up leaving the lab, too.

Unless there is a documentary trail of falsification, people like him can just keep on committing fraud without consequences. If I had evidence, I would speak out.

In my freely available autobiography, I describe an incident where I also came under pressure to violate the scientific rules, which happened when I was a junior doctor, only a couple of years after my graduation.

To avoid bias during data analysis and the writing of manuscripts for clinical trials, I had introduced blinding during curation and analysis of the data, and I even wrote the manuscripts blindly. This was so unique that I was invited to give a lecture about it, with all costs paid, at the annual meeting for the Society of Clinical Trials in Houston in 1994, and I published an article about my methods and practical experiences with them.

According to my procedure, I wrote two versions of the manuscript. It was particularly interesting to write two discussion sections, not knowing what was A and B, and I required of my co-authors that they approve both versions before we broke the blinding.

Because of my profound knowledge of statistics and trial methodology, I had been invited to be a co-author when colleagues at the Department of Infectious Diseases at Rigshospitalet in Denmark planned a short-term, placebo-controlled pilot study of the effect of fusidic acid – an antibiotic - in African AIDS patients.

I did the data analysis, and there was a mortality difference. The P-value was 0.06 with Fisher’s exact test and 0.02 in a survival analysis, week by week. Despite the randomisation, there was a marked difference in weight at baseline, which was important because weight was strongly correlated with survival. Adjusting for weight yielded a P-value of 0.13.

This was an exciting moment, with P-values spanning from 0.02 to 0.13. In such situations, many researchers let their bias determine what they choose to report and how.

In vitro studies had suggested that the drug had an effect against HIV. Had we shown that this effect could be replicated in patients? Or did the drug have a deleterious effect on weak AIDS patients that killed some of them? Or didn’t the drug do anything?

During the discussions of the two blinded manuscripts, my co-authors suggested breaking the code to guide the analysis, but I refused. Under the null hypothesis of no difference in effect between drug and placebo, any decision taken during data analysis should be independent of the treatment code. Furthermore, I had specified the principle of blinding during data analysis and writing of the manuscript in the trial protocol.

After both manuscripts had been accepted by my co-authors and the code was broken, I was again put under pressure to make statistical adjustments, for other imbalances than weight, but now my colleagues knew in what direction any adjustment would change the result. When I refused, one of my co-authors produced a new manuscript with new analyses, some of which were not only unwarranted but illegitimate, e.g. he had excluded some deaths that occurred after randomisation, which is not allowed.

There were heated disputes over this. I needed to tell my co-authors that if they went ahead with their inappropriate analyses, I would withdraw as author and would publish a letter to the editor where they published the trial explaining that they had violated the protocol. This calmed them down and convinced them to adhere to our protocol.

The end of the story was that a second, larger study failed to confirm the nonsignificant tendency in the pilot trial. This trial was not published.

I was not involved with the larger study. I would have insisted that the results should be published. Not to publish negative results is unethical and scientific misconduct. And yet, we cannot even know if what gets published is truthful. What really happens in a trial is often not reflected in what is published, but the readers rarely have any chance to investigate if the authors committed fraud, which often happens, even in our most prestigious medical journals.

I surely ran a risk. I went directly against my professor, as it was his idea that the drug might work against AIDS. When I told the story to colleagues, they were shocked that I dared do what I did, as professors can ruin careers. Many people who have made it to the top have psychopathic traits but, in this case, the professor was an honourable person and he appreciated my expertise. I could not know if my stubbornness would have made him my enemy.

Healthcare is hugely corrupted by industry money and too many doctors have a price at which they can be bought. I have never considered truth negotiable. Together with freedom of speech, it is the most precious asset we have.

I have never feared for the consequences for myself when I stood my ground. If we don’t react and stop psychopaths, we will go from bad to worse. It is not an option to stay quiet when we see fraud or injustice.

Desmond Tutu, Archbishop and Nobel Peace Prize Winner from South Africa, said: ”If you are neutral in situations of injustice, you have chosen the side of the oppressor.” He also said: “If an elephant has its foot on the tail of a mouse and you say that you are neutral, the mouse will not appreciate your neutrality.”