Once enrollment and data collecting stop, then what?

Determining when to close the file

As long as local protocols continue to enroll subjects and collect data, they must submit to annual reviews by the institutional review board. But what happens once the study concludes, enrollment is closed, and all data are collected? What obligation does the IRB have to continue monitoring the study?

According to the Office for Human Research Protection (OHRP), protocols must continue to be reviewed annually while any data are being collected or analyzed or any interaction with human subjects continues. However, studies are eligible for expedited annual review once the following three situations occur:

• (i) the research is permanently closed to the enrollment of new subjects; (ii) all subjects have completed all research-related interventions; and (iii) the research remains active only for long-term follow-up of subjects;

• no subjects have been enrolled and no additional risks have been identified;

• the remaining research activities are limited to data analysis.

The major benefit is that expedited reviews can be conducted by a smaller group of the local IRB and not by a convened meeting of the entire IRB, says Don Workman, CIP, director of the Office for Protection of Research Subjects at the University of Illinois at Chicago (UIC).

"A convened meeting must be attended by a majority of the IRB members, including at least one nonscientific person," he explains. "An expedited review can be conducted by the IRB chair or by one or more experienced reviewers designated by the chair from among the members."

Though a less formal review, the OHRP has been clear that they expect to see a thorough examination of the study design, the questions the study is attempting to answer and what progress has been made, the number of subjects enrolled, the adverse events reported, and how those risks have been addressed and minimized.

The IRB also must have an established protocol for how to communicate information about the studies receiving expedited review to the entire board, and must demonstrate that the review process is not just a cursory glance, he adds.

"A lot of IRBs have gotten into trouble by lumping 10 protocols up for continuing review and taking one motion, something like saying, Does anyone have an issue with any of the studies up for continuing review?’" he explains. "You shouldn’t have these blanket approvals. You need to review each one individually."

The procedures for reporting adverse events throughout the trial and the procedures for annual review both serve as feedback loops for the IRB to monitor the study’s progress, with the adverse events serving as a sort of daily check and the annual review as a time for the IRB to comprehensively look at all of the information gathered over the course of the previous year, Workman says.

The smaller subset of the IRB that conducts an expedited review is not able to disapprove a study — that can only be done in a convened meeting — but they can require the study to be suspended while the investigators address any critical issues found.

That requirement prevents situations in which one person might be able to hurt another’s research, but allows problems to be addressed pending the review of the entire board.

When it’s finally over

But at what point can the IRB stop conducting annual reviews?

That question is not explicitly addressed in the regulations, says Workman, but it is generally accepted that a study is concluded once all interaction with study subjects has ceased, and all data have been collected and analyzed.

At that point, continuing review no longer is needed.

However, there is no broad consensus about what constitutes proper analysis of the data, Workman says.

Investigators usually prepare some sort of final report to the IRB indicating that the study has concluded and what the results were, he notes. "Otherwise, you would just end up with these protocols with lapsed approvals, and you need to know what the status is."

However, some experts feel that IRBs should continue review until there is some sort of publication of the study results or the information is made public in some way, Workman adds.

Human subjects agree to participate in research with the knowledge that the information will be put to use. IRBs must be careful to not support or encourage investigators to abandon data, particularly if it happens to be data that the study sponsor is just not happy with, he notes.

There is a great concern among research ethicists now about pressures, both real and potential, on investigators to not publish results that are unfavorable to a sponsor’s test drug or product.

"Sometimes you just have a study and the results are just not what the investigators expected, or they just don’t reveal anything in particular," Workman notes. But there are journals that publish null results — and even that type of publication is useful in that it prevents other researchers from duplicating a bad study. "That’s not what everyone wants. There’s not a lot of incentive for journals to publish the results of a study that went nowhere, but there is value in it."

At St. Jude Children’s Research Hospital, where Workman worked prior to UIC, a group of investigators started a small-scale study whose design was picked up by the Children’s Oncology Group, which expanded it, he notes.

The original researchers then closed their study, but the local IRB did not want to abandon the information collected from the participants there.

"They really worked with the investigators and kept them coming back until they found some forum to publish the data," he notes. "Now you are talking about uncharted waters, because there is certainly nothing in the regs that says an IRB can keep a study under review until the investigators find a journal willing to publish, but because of the ethical concerns, they and the investigators were willing to."