Markets

A Glass Half Full: Avandia Case Reveals Hole in FDA's Conflict-of-Interest Rules

By Jim Edwards

Updated on: July 20, 2010 / 12:39 PM EDT / MoneyWatch

The news that one of the doctors on an FDA panel assessing whether GlaxoSmithKline (GSK)'s diabetes drug Avandia causes too many heart attacks failed to disclose he was a paid speaker for the company points out a giant hole in the FDA's regulations: The disclosure form that outside experts who advise the FDA on risky drugs are required to sign only requires experts to list fees from speaking or writing for a drug company for the "Last 12 months or under negotiation." That's too short a time period to catch most conflicts in the drug business.

The Wall Street Journal reported that endocrinologist David Capuzzi -- one of only three experts who voted for Avandia to stay on the market with no additional warnings -- received $14,750 in total from GSK over the last couple of years. (His lecture topic was Lovaza, a heart drug, not Avandia.)

But if you check GSK's payment disclosure forms back through Q2 2009 (and GSK's admission that it paid Capuzzi $3,000 in Q2 2010 and another $8,000 before it started disclosing the payments publicly) it turns out that even if Capuzzi had filled in the FDA form correctly he only would have had to report $6,750 of his GSK income -- less than half his actual financial conflict with the company.

UPDATE: The same thing applies to Dr. Abraham Thomas, who voted to remove Avandia. He was a speaker between September 2007 and September 2008 for Takeda. If he had filled in the form correctly his conflict would not be noted because it occurred more than 12 months ago.

Clearly, 12 months is too short a timeframe to accurately judge whether someone is conflicted or not. Research work sometimes takes years to complete, and publishing studies afterwards can add even more years to the process. BNET readers discovered how a too-short timeframe can cover up potential conflicts when I noted that Washington University School of Medicine's Dr. Joan Luby, who authored a paper recommending that 3-year-olds be given antipsychotics, had failed to disclose in 2009 that she received payments from Janssen (the unit of Johnson & Johnson that markets Risperdal), or that she has given talks sponsored by AstraZeneca (maker of Seroquel), and has been a consultant for Shire (maker of Adderall XR and Vyvanse).

Some medical journals require conflict disclosures going back as far as five years. The FDA should do the same.

Related:

Image by Flickr user Todd Porter, CC.