Review warning letters to prepare for audits

Letters can show what FDA is looking for

A survey of warning letters sent to IRBs by the FDA shows that most of the offenders were deficient in some of the most fundamental areas of responsibility — failing to have or follow adequate review procedures and failure to maintain adequate documentation of IRB activities.

The survey, published in the December issue of the journal Clinical and Investigative Medicine, looked at 7½ years’ worth of FDA warning letters. It found that most of the 52 IRBs that received them were warned about more than one violation.

While the total number of IRBs who were issued such letters could be seen as small, the authors argue that their contents "consistently indicate weaknesses in the review and documentation activities of audited IRBs, potentially signaling similar issues among all IRBs across the United States."

The study notes that it reported only the violations reported by the FDA, and did not attempt to investigate the validity of the findings.

One of the study’s authors notes that examining such warning letters can be a good way to educate IRB members who’ve never been audited as to what the FDA is looking for in audits and to alert them to areas that should be carefully monitored within their own institutions.

"If they saw some of these letters — not that they don’t take their job seriously already — but the point would really be driven home on how serious their job is and how seriously the FDA takes their job," says Katrina Bramstedt, PhD, bioethicist and director of the General Clinical Research Center’s Research Subject Advocacy Program at the Cleveland Clinic Foundation.

"This wasn’t necessarily designed to poke fingers at poorly performing IRBs per se, but really it’s an educational piece, asking, How can IRBs be helped?’"

The study grew out of a similar review of FDA warning letters to clinical investigators, Bramstedt says. She says she uses such letters as part of her education of principal investigators about research ethics.

"If you just tell them all about the [Code of Federal Regulations] and all the rules and the regs, it’s really boring, and it just kind of rolls off the top of their heads," Bramstedt says. "But if you show them warning letters, they really sit up and take notice. Then they see, wow, this stuff is real. It’s not just a law in a book, this is what happens when these people come in and audit you."

Most issued to hospitals

FDA warning letters are posted on the agency’s web site at www.fda.gov/foi/warning.htm. As Bramstedt reviewed warning letters sent to clinical investigators for her previous study, she also noticed letters being sent to IRBs, and decided to examine them as well. She searched the warning letters database by using the terms "IRB" and "institutional review board."

According to a 1998 report by the Office of the Inspector General, the FDA conducts about 200 audits a year of IRBs. The purpose of the audits is to determine whether IRBs are following both federal regulations and their own internal operating procedures.

In reviewing warning letters from January 1997 through July 2004, Bramstedt found 52 issued to IRBs. Of that number, 34, or 65%, were issued to medical centers or hospitals, compared to nine each (or 17% each) sent to university IRBs and private IRBs.

Bramstedt says she’s unsure why medical centers and hospitals received so many more warning letters.

"It could be that there are simply more medical center IRBs than university ones," she says. "It could be that more medical centers are audited compared to the others. It’s hard to know because you don’t have the baseline to go by. It would just be guessing."

The most common regulatory violation reported in the warning letters was the failure to have and follow adequate written procedures on how the IRB is to conduct reviews of research — 50 of the 52 letters, or 96%, cited this violation. The second most common, failure to prepare and maintain adequate documentation of IRB activities, affected nearly as many IRBs — 47, or 90%.

The study argues that the violations alleged by the FDA are not merely perfunctory recordkeeping issues.

Having and following adequate procedures ensures that the projects are all reviewed in the same way, the study notes. This is important not just to the scientific integrity of the research, but also to safeguard the health of participants, in particular the treatment of vulnerable populations.

As for the documentation requirements, the study notes: "This is serious. Such records are pivotal in cases of research controversy [e.g., who did/ did not vote, the vote results, research changes requested and their rationale]."

Other violations reported in the warning letters included failure to provide adequate initial review and continuing review of studies (11 letters and 36 letters, respectively); failure to meet membership and/or quorum requirements, including conflict of interest for IRB members (30 letters); and failure to meet federal guidelines in the informed consent process (19 letters).

Bramstedt says informed consent issues were also a theme in the letters sent to clinical investigators. "Consent forms that were written too complexly, consent forms that might have hyper-inflated benefits to subjects, or failed to include pertinent risks, things like that," she explains.

Letters raised cultural concerns

While there was not a lot of detailed information about violations in many of the letters Bramstedt reviewed, she says two in particular caught her interest because of the cultural issues they raised.

One involved a university IRB that had approved a study involving direct injection of blood from one person to another and the administration of live malaria parasites to research subjects.

"The FDA said there was no documentation that indicated the IRB had reflected on how these two practices might be perceived in light of cultural attitudes in overseas research subjects," she says.

The other case involved a private IRB doing research in Puerto Rico. The IRB had approved two clinical trials conducted on research subjects who were potentially "economically disadvantaged."

The warning letter states that during the IRB’s review, "nobody ever explored that whole area of using a vulnerable population like that," Bramstedt says.

Seventeen letters, or 33%, of the total letters reviewed included at least one restriction imposed upon the IRB by the FDA. Restrictions included ordering IRBs not to review or approve any new studies; ordering that no new subjects be enrolled in ongoing studies; and in one case, ordering the IRB chair to immediately contact the FDA to arrange a meeting to discuss corrective action.

The authors noted that one limitation of their study was that they didn’t have access to all follow-up letters from the IRBs, which would give a better idea of the outcome of the FDA’s actions. Although IRBs can post their response letters on the FDA’s web site, it is not required, and only one letter was posted.

Bramstedt says the number of warning letters varied widely over the years she studied, ranging from a high of 12 in 1999 to a low of two in 2001.

She says she actually expected to find more warning letters, given the number of audits conducted.

"But then again, if they’re significant findings, then that’s potentially relative to research subjects’ safety, which of course is the important thing," she says. "It’s not necessarily the findings themselves, but do they translate to problems with research subjects’ safety? It’s an extrapolation, but one could venture that yes, there’s probably a connection."

She reiterated that her goal in this study wasn’t to find fault, but to alert IRBs about what FDA audits at other institutions have found.

"I think it’s good to have a survey of these warning letters because if you’ve never been audited by the FDA, you might not know what to expect," Bramstedt says. "This way you know what they’re looking for, you know their hot buttons, the things that are going to be key.

"So I think that’s really important. I think it can help you prepare for an audit."