Healthy Skepticism Library item: 10939

Warning: This library includes all items relevant to health product marketing that we are aware of regardless of quality. Often we do not agree with all or part of the contents.

Publication type: news

Silverman E.
Cancerous Data? Novartis Statistics Whiz Blows The Whistle On Data Reporting System
Pharmalot 2007 Jul 19
http://www.pharmalot.com/2007/07/cancerous-data-novartis-stats-whiz-blows-whistle/

Notes:

Link to lawsuit: http://www.pharmalot.com/wp-content/uploads/2007/07/tasigna.pdf

Reuters report on Tasigna approval delay: http://www.reuters.com/article/businessNews/idUSL1718021220070717?pageNumber=1

Full text:

As David Olagunju tells it, he enjoyed his work at Novartis, where he was global director of biostatistics and statistical reporting standards, until early last year. That’s when he was shifted to the oncology division, and was soon reporting to a new supervisor. The rest of 2006, he says, was filled with nothing but anxiety. In fact, by this past January, he was out of a job.

Why? In a lawsuit filed in May in a New Jersey state court, the 59-year-old Nigerian native alleges he uncovered numerous violations of the drugmaker’s safety reporting system, which is used to process safety data from clinical trials, in particular the Tasigna cancer med trials. He reported the problems internally, but claims to have suffered only retaliation, such as negative performance reports. A Novartis spokeswoman declined to comment. Here’s the lawsuit.

Since then, Olagunju says he met last month with FDA inspectors and gave them documents – protocols, statistical analyses, e-mails – to support his claims. An FDA spokeswoman would only say the agency “doesn’t have any information to provide.” Olagunju, however, is asking the FDA to audit the Tasigna trials. Coincidentally, Novartis disclosed this week that Tasigna approval was delayed three months while the FDA reviews additional data; no new studies were required.

Among the allegations: The reporting system wasn’t validated according to regulatory standards and generated errors – empty files, poor documentation, incorrect date imputation for adverse events, wrong variables, and mismatched items. In fact, Olagunju says the problems surfaced in 2004, but worsened in early 2006, prompting him to complain. Ultimately, he adds, this allegedy broken system was used to crunch data for 148 different clinical trials.

Another example: a contract programmer complained of being forced to alter the calculation of the duration of adverse events to enhance the appearance of Tasigna’s safety in clinical trials. Then, the Development Leadership Team, which is a group of vice presidents, allegedly redefined the details of how patients were randomly assigned to treatment drugs, and how the effect of the drugs were measured, reported statistically and negotiated with the FDA a few days prior to locking the data base.

The bottom line: “By compressing the duration of adverse effects, Novartis has exposed the public to a serious danger if this drug were to be approved…These issues have serious impact on the reporting of safety data and have negative impact all of Novartis’ clinical trials, not just the clinical trials for Tasigna,” according to the lawsuit. Again, however, these are only allegations and the FDA isn’t saying anything one way or the other.

“I followed the company’s policy of reporting such problems to the letter. I went step by step. But still they did nothing and kept playing games,” he tells Pharmalot. “This was the last step anyone wants to take. But I pointed it out to my supervisor and I was snubbed. I went to HR and was snubbed. And I went to corporate compliance and I was snubbed. So there’s no choice. You have to defend yourself and tell the story. But I would say that unless you have money and courage, keep your mouth shut. Or you should probably just go to the authorities to report it.”

Peter Rost at Question Authority has also written [on] this.