The trusted source for
healthcare information and
Many IRBs may not know member interest conflicts
Talk to your institution first, to see how others report
A significant minority of IRBs at some of the nation's biggest medical institutions lack sufficient procedures to determine when IRB members have industry relationships that could pose a conflict in their work.
That's the assessment of researchers who surveyed IRB chairs at the United States' most research-intensive medical institutions. Their findings were published earlier this year in the journal Academic Medicine. 1
The findings showed that a third of IRBs represented in the 2005 survey had no requirement that voting members disclose relationships with industry. A quarter of the IRBs didn't have written policies outlining what to do if a member's conflict of interest was identified.
Twenty percent of IRB chairs surveyed said they didn't feel confident that their policies and procedures ensured adequate disclosure of members' industry ties.
Previous research has shown that some IRB members have had such industry relationships — research funding, work as paid consultants or as members of advisory boards, royalty earnings, or serving as officers of companies.
Christine Vogeli, PhD, instructor of medicine at Massachusetts General Hospital and Harvard Medical School in Boston, MA, and the lead author of the study, says that while the original survey was administered in 2005, she doesn't believe the situation has changed significantly since then.
"A lot of information (about conflicts of interest) has come out in terms of institutions in general, but not specifically related to IRBs," Vogeli says. "Although institutions may have set up practices to be able to record member relationships, it's not clear that those extend to the IRBs."
She says she was surprised at the results, given that federal guidance and professional associations strongly recommend that IRBs develop such policies and procedures.
"These are some of the top institutions, the ones that really bring in most of the money and do most of the research within this country," Vogeli says. "So I guess I would have hoped that there would have been a little bit more consistency in application of the guidance."
A total of 107 active IRB chairs responded to a mailing to IRBs at the 100 medical schools and 15 independent hospitals that received the most funding from the National Institutes of Health in 2003. They replied anonymously to a series of questions about the processes in place at their institutions to manage member relationships with industry.
Of that group, 70 (or 66%) reported that voting members were required to disclose such relationships, and 25 (23.6%) required non-voting members to do so. A total of 79 chairs (74.5%) reported that their IRBs had a defined process for disclosing members' industry relationships.
There was wide variance in how members reported relationships — some reported to the chair, some to the entire IRB and others to another group or individual outside the IRB, such as a conflict of interest committee or a senior administration official.
Similarly, of the IRB chairs who identified a conflict of interest arising in the previous year, 31% said that the decision about how to handle it was made by the entire IRB, 27% reported the IRB chair made the decision, 16% said it was left to a group outside the IRB and 13% said the individual IRB member affected made the decision.
Comparisons of this survey of IRB chairs to a companion survey conducted of IRB members from the same group of institutions turned up a significant discrepancy. All 47 chairs (100%) who had reported a conflict of interest in the prior year said their members never voted on protocols on which they had a conflict. However, in the companion survey, only 64.5% of members with a conflict said they never voted on the protocols in question — 19.4% said they always voted on protocols on which they had a conflict of interest.
While the anonymity of both surveys made it impossible to know which members' responses might match up with which chairs' results, Vogeli says the point is still telling.
"I think the IRBs think they know about the relationships with their members and the conflicts that they have, but they really don't," she says. "It shows there's definitely a gap. A third of members reported that they had at some point voted on a protocol in which they had a conflict, while the IRB chairs thought that no one had. That's problematic."
Even some of the IRB chairs surveyed showed concern over the issue. Twenty% of them reported that they were "not very confident" or "not at all confident" that their procedures were providing the appropriate level of disclosure for member-industry relationships.
Looking to the institution
While Vogeli says it's important that IRBs develop policies for recording and responding to members' potential conflicts of interest, she doesn't think the answer is for the IRB to simply come up with its own plan.
Instead, she says, the first step should be for the IRB to look to its institution's existing policies and to consider an approach that would mesh them.
"I think it requires a conversation between the IRB and the institutional leaders to say, 'What are you doing? Is it sufficient for what we need to do at the IRB, should we supplement that or can we base our findings of relationships on your data collection?'"
She says there's no one approach that necessarily should apply to every IRB or every institution.
"I think there are a number of different ways this can be handled," Vogeli says "But you need to handle it."
The potential implications of undisclosed conflicts of interest add a sense of urgency to the issue, Vogeli says. She says an IRB doesn't want to approval a protocol on the advice of someone who had an undisclosed relationship and a subject is later injured or dies.
"You can think of a scenario where you would say, it would have been very good if we had known that he had a relationship," she says. "It might have changed the types of questions we would have asked in our review, or the attention that the other members would have paid to it.
"I just think knowledge really helps drive what happens in an IRB, where everybody has the best intentions but you might not ask the questions you would ask if you really knew."