Honest scientists can recognize inappropriate research behaviors when they see them. Consider the following:
This was the portion of a 1983 Pfizer document that reported that one patient had been taken off Zoloft (an antidepressant) during a clinical trial. The portion of the report that had been in the public domain (and presumably also with the FDA) from 1983 to 2004 indicated only that the patient was taken off the drug because of treatment failure and nausea, anorexia and painful urination.
The true clinical trial record indicated that the patient had been withdrawn at day 11 of the trial because of thoughts of killing himself and others : "[The patient] began to verbalize feelings of killing other people and then himself."
Anyone who cares at all about honest medicine will understand the significance of the subtle redrafting of this and many other records in clinical trials. See original documents here. Read also here for details of the court case that led to release of this particular record.
Why should companies be allowed to maintain the records of clinical trials when human lives and so much money are at stake, and where there have been so many instances of misconduct? The design of a research protocol or system that allows for this form of corporate control is in my view scientific misconduct in and of itself.
And the response from Pfizer at the time: "This is anecdotal patient information," said Bryant Haskins from the company's New York office. A clinical trial is the systematic collection and honest statistical description of anecdote. When each bit of that information is distorted, what we end up with has the mere appearance of science. In medicine, research misconduct leads to death, incorrect decision making, ineffective therapy and misery.
Has anyone been punished for the misery caused? We feel smug self-satisfaction when we send an individual scientist such as Eric Poehlman to prison for scientific misconduct while we ignore even the most severe instances of corporate scientific misconduct.
See also Pharmagossip: Those who forget history for a discussion of the retrospective massaging of one individual patient event in Merck Vioxx trials.
Much of what we call "science" in pharmaceutical medicine doesn't pass the most basic sniff test.
Take the survey. Do these instances represent scientific misconduct? If you answer no, please leave a comment justifying your opinion.