Members of drug advisory committees at the US Food and Drug Administration often have financial conflicts of interest and those conflicts affect voting patterns, says a study in JAMA (2006;295: 192116639051).
In 73% of the 221 meetings analysed, at least one advisory member or consultant had one or more conflicts. On an individual level, 28% of advisory members and voting consultants had conflicts. The researchers found that if panellists with conflicts had been excluded, voting margins for the index drug would have been less favourable. In none of the instances studied would exclusions have changed the majority vote for or against approval.
The study, by Peter Lurie of the Public Citizen's Health Research Group in Washington, DC, and colleagues, analysed data from agency transcripts over a four year period, from 1 January 2001 to 31 December 2004. The most common conflicts of interest were consulting arrangements, contracts or grants, and investments. Nineteen per cent of consultancies involved payments over $10 000 (£5600; €8100); 23% of contracts or grants were over $100 000; 30% of investments were over $25 000; and 44% of lecturing honorariums were over $10 000. Despite the frequency of conflicts, only 1% of panellists with conflicts were “recused”—that is, disqualified because of their personal involvement.
The researchers studied three categories of conflicts of interests: “index conflicts” (those with financial interest in the drug under discussion); “competitor conflicts,” and “any conflict.” In all three categories, voting margins would have been less favourable to the index drug if panellists with conflicts had been excluded. Panellists with competitor conflicts sometimes exhibited the widest margin in favour of a drug, something the authors suggested might be due to a general “pro-industry” position among those with any industry ties.
The impact of the agency's January 2002 draft guidance on conflicts of interest was mixed, say the researchers. Compliance with provision of details about the types of conflicts and their monetary values approached 100% after the guidance. However, the guidance did not require the name of the competitor company, and disclosure of this information dropped from 54% to 1%.
In at least 32 cases, patient advocacy groups that were funded by a drug company provided speakers during the public sessions, and in 47 instances a public speaker was flown in by a drug sponsor. This, said the researchers, “amplifies the growing concern that pharmaceutical industry sponsorship is becoming more prominent in nonprofit, patient advocacy groups that were once viewed as grassroots organizations independent of industry influence.”
The researchers found that both individual level and meeting level effects were created by conflicts of interest. Meeting level effects can arise when panellists are affected by the votes of others or by the presentation of public speakers—swaying the entire panel. The authors point out that the recent advisory panel's review of COX 2 (cyclo-oxygenase-2) inhibitors, which occurred outside of the time frame of the current analysis, would have had a different outcome in two of the votes taken if panellists with conflicts had been excluded.
Dr Lurie told the BMJ, “Our tolerance for undue influence at a committee as important as an FDA advisory committee should be zero.”