Averting low-enrolling studies benefits IRBs
Studies that fail to enroll sap IRB resources, raise ethical issues
When a study is terminated because of low enrollment, it wastes the institutional resources that allowed it to be started — including those of the IRB that approved it.
So identifying factors that can lead to low enrollment and weeding out studies unlikely to be completed makes good business sense for an institution. It can also make good ethical sense, since approving a study that can't be completed produces no scientific benefit to justify the risks of the few subjects who are enrolled.
To get at why some studies don't enroll subjects, investigators at Oregon Health and Science University analyzed 260 clinical studies that were terminated at OHSU over a three-year period because of low enrollment. They looked at factors such as type of IRB review and type of funding to see which studies were more likely not to sign up participants.
They also calculated the cost of those studies to the institution — nearly $1 million for fiscal year 2009.
"That's a very conservative estimate," says Darlene Kitterman, MBA, director of Investigative Support and Integration Services at the Oregon Clinical and Translational Research Institute in Portland. "That's just what we could quantify. There are a lot of non-quantifiables that we list in the article that we couldn't quantify in a reliable way."
IRB costs accounted for more than $100,000 of that amount — including IRB preparation, the costs of full board reviews and continuing reviews.
Kitterman says she could see the effects of these studies on her institution's resources.
"And I would even hear from investigators sometimes that they really didn't think they'd enroll anyone, but they wanted the study to get started anyway," she says. "The perception that that was a good thing was an opinion that I wanted to sway people from having."
Electronic tracking
Kitterman says that her institution was an early adopter of electronic systems that track studies, which made it easier to calculate the costs of these low-enrolling studies.
"A lot of this information would be really hard to gather if it were not in a database somewhere." she says.
Her team chose to look at studies that enrolled only zero or one participant by the time of termination. "When I started looking at the data, there were high enough numbers of zero-to-one that I thought that would be shocking enough. It gave the message."
The 260 studies terminated for low enrollment represented nearly a third of all terminated studies during fiscal years 2006-2009.
Of that low-enrolling group, 173 studies, or 67 percent, went through full IRB review.
Kitterman says she believes that points to study complexity being a factor in low-enrolling studies.
"A lot of these studies fail in feasibility and specifically, operational feasibility," she says. "The more complicated the study is, the more likely that is to happen. Maybe the eligibility criteria are not realistic. Or it's using supplies or equipment that we actually don't have at OHSU, but that wasn't considered before the study was implemented.
"Or maybe those kinds of patients really don't exist here and we didn't realize that because we didn't do a very good recruitment plan ahead of time. The more complicated a study is, the more likely it is to fail out from an enrollment perspective."
Investigators also identified government-funded studies as more problematic. Of the 97 total government-funded studies terminated during the three-year period, 53 percent were low-enrolling (compared to 38 percent of all industry-sponsored studies terminated during the same period).
"A lot of those are things like cooperative group studies, where there's a lot of pressure to participate even if you're not going to enroll," Kitterman says.
Absent PIs
The team chose a small group of studies and asked investigators directly about the reasons their studies didn't enroll subjects.
One reason that came to light: Often these studies were started by investigators who were preparing to leave the institution, but opened the study anyway.
"They would transfer them to another PI, and subsequently they didn't enroll," Kitterman says. "The hypothesis being that the study was just sort of stuck with someone who wasn't necessarily interested in it."
Having identified some areas to focus on, Kitterman says her institution has begun educating investigators and department heads about the costs of low-enrolling studies and the importance of determining the feasibility of a study before moving forward with it.
"I think part of it is just having an acknowledgement of the issue," she says. "That you need to do a more thorough feasibility review, and that if you know you're not going to enroll, or it's a good possibility that you're not going to enroll, maybe you shouldn't do the study. That's pretty simple, that's really education and perhaps some oversight."
The group is continuing to monitor low-enrolling studies to see if this educational intervention works. Kitterman says there are other options the institution could employ, such as requiring a formal feasibility review of studies.
She says that while IRBs should withhold approval of a study if they find it unfeasible, ideally, an unfeasible study shouldn't make it that far in the process.
"Reviewers have clinical expertise in that protocol, but that doesn't mean they necessarily have operational expertise," Kitterman says. "I would really like to see these kinds of studies not get to the IRB. That's what I would think a successful intervention is."
Reference
Kitterman DR, Chang SK, Dilts DM, et al. The Prevalence and Economic Impact of Low-Enrolling Clinical Studies at an Academic Medical Center. Acad Med 2011 Nov;86(11):1360-1366.
When a study is terminated because of low enrollment, it wastes the institutional resources that allowed it to be started including those of the IRB that approved it.You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content