Report Finds “Conflicts of Interest” Have No Effect on FDA Advisory Committee Votes
Stringent conflicts-of-interest policies keep many experts off of FDA advisory committees. A new study suggests that the fear of pro-industry bias underlying these policies may be misplaced, and also serves to keep highly qualified candidates off of these committees.
James C. Cooper, director of research and policy at the Law and Economics Center at George Mason Law School and Joseph Golec, professor of Finance of the University of Connecticut, who conducted the study, sought to compare conflicted members’ voting patterns with objective criteria. They found that decisions by advisory committees with conflicted members to recommend drugs were more likely to be consistent with both the ultimate FDA decision as well as stock market predictions than non-conflicted advisory committees and members.
Background
In the past few years, there has been an increased focus from both the media and responsive legislation on financial ties between advisory committee experts and drug companies.
Cooper and Golec’s study notes that the law casts a wide net on what is a potential conflict of interest, including a situation where an oncology researcher would be “conflicted” if the university where she works received funds from the sponsor drug company, even if none of it funded general oncology research or her specific research. Furthermore, the FDA Amendments Act of 2007 placed additional restrictions on advisory committees, requiring that no more than 13 percent of member participants per year could receive conflict waivers to allow them to participate in a meeting. The study notes that FDA also reduced the maximum size of financial interests eligible for waivers from a combined financial interest of up to $100,000, to a maximum of $50,000.
The reasoning behind these policies is that financial ties between advisory committee experts and drug companies could potentially bias committee members’ recommendations in favor of unsafe or ineffective drugs. However, the authors point out that the “same specialized education and scientific experience that makes these experts attractive candidates to serve on [advisory committees] also puts them in contact with drug companies.”
Pharmaceutical companies often employ or contact these same expert physicians, researchers, and clinicians to help them develop and potentially market products. “For instance, companies seek out these experts to monitor or run their clinical trials, speak at various company-sponsored meetings, consult, write briefs, or serve on review boards,” states the report. “Drug companies, moreover, fund research studies at universities and research institutes that employ these experts.”
Because of the expertise these conflicted experts possess, restricting their participation on advisory committees is not a perfect solution, the authors argue. Limiting the number of conflicted experts who can serve on advisory committees presents its own costs associated with finding qualified, non-conflicted members and appointing less-qualified members.
Study Findings
The authors’ main goal in conducting their study was to see if the restrictive policies for FDA advisory committees in fact brought unbiased outcomes. “If such financial ties lead to FDA approval of unsafe or ineffective drugs, then the costs may be justified,” they note. If these ties have no impact on approval decisions, however, the authors believe the rules that place excessive barriers to entry can do more harm than good. Indeed, “such policies can impose net costs on society, especially if less expert committee members misinterpret the scientific data and recommend approval of unsafe or ineffective drugs.”
Cooper and Golec conducted a previous study in 2013 which found “no statistically significant difference between conflicted and non-conflicted member voting.” To conduct this follow-up 2015 report, the authors examined the extent to which the presence of conflicts impacted advisory committee decisions as compared to two benchmarks–the ultimate FDA decisions and stock market predictions.
As for the first benchmark, the authors found FDA’s ultimate decision to be a useful barometer because FDA staff are often hired specifically for their scientific backgrounds and expertise in evaluating the safety and effectiveness of new drugs. Furthermore, FDA experts evaluate hundreds of drugs each year, and are “employed directly by the entity that is politically accountable for poor decision-making.” The authors also point out that FDA’s noted history of “risk-aversion” suggests a bias away from drug approval.
The study also benchmarks its committee decisions with decisions from investment companies. These companies hire experts to evaluate the scientific information supporting potential new drugs, and advise their clients about the likelihood of advisory committee and FDA approval, states the report. An advisory committee decision that “surprises the market – as reflected in abnormal returns measured through event studies – can be classified as ‘incorrect’ in the sense that it is contrary to the best judgment of outside experts.”
The results of the benchmarking study were that conflicted advisory committee members, which the authors abbreviate “AC,” were potentially more consistent with unbiased benchmarkers than their non-conflicted fellow members:
“The proportion of conflicted members serving on an AC has no statistically measurable impact on the probability that an AC decision is contrary to that of the FDA. The presence of conflicted members, moreover, reduces the probability of disagreement with the FDA for approval decisions. The empirical results thus suggest that if anything, conflicted ACs are more likely than their non-conflicted counterparts to disagree with the FDA’s decision to approve a drug. Results using the stock market as a benchmark are similar: conflicted ACs’ decisions surprise the stock market less frequently and are associated with smaller average abnormal returns than those made by their nonconflicted counterparts. These results suggest that conflicted ACs tend to arrive at decisions that are more consistent with predictions by outside experts than non-conflicted ACs, perhaps reflecting their greater expertise.”
—
While the George Mason University study may not change the mind of the most hardened industry critic, it is not the first study to undermine the conflict-of-interest trend. The FDA commissioned a study in 2009 entitled Financial Conflict-of-Interest Disclosure and Voting Patterns at FDA Advisory Committee Meetings that “found no evidence to suggest that having a financial conflict-of-interest tends to increase votes in favor of that interest.”