Finding ethical balance on protocol violations
A real world problem
There isn't a black and white answer when an IRB discovers that a human subjects research site has collected some data that is tainted by a protocol violation.
"This is a question of degree," says Bruce Levin, PhD, professor and chair in the department of biostatistics at Columbia University's Mailman School of Public Health in New York, NY. Levin spoke about data collected out of compliance at the Columbia University Institutional Review Board Educational Conference, held March 17, 2009, in New York.
It's unrealistic for IRBs to insist that clinical trial sites perform with no protocol violations, Levin notes.
"The real world of doing clinical studies has the ideal of following the protocol, but it's not often possible," Levin says. "Some would say it's not ever possible to follow it exactly to the letter, so the question is 'What is the magnitude, the frequency of protocol violations?'"
Most phase III, randomized, large clinical trials have a mechanism for reporting protocol violations, and most of the time they're not serious and nothing happens as a result of it, he explains.
"Sometimes they're serious and have to be reported to the IRB and data safety monitoring board and regulatory agencies, as well," Levin says. "So it's not easy to give a black and white answer."
One of the ethical challenges is that how an IRB should handle this might depend in some part on how large the study is.
"If one is focusing on small scale studies, the issues become potentially more serious," Levin says.
For instance, if a small study with 20 enrolled participants has a protocol violation that could result in data from four or five subjects being excluded in the analysis, then this would mean throwing out 20 to 25 percent of the sample, Levin says.
"This would be a serious degree of missing data," he adds.
"Of course the dilemma is when you include data that is seriously out of compliance you run the risk of potentially not knowing what experiment you're conducting," Levin says. "The world will understand the experiment according to the protocol."
As a biostatistician, Levin has worked with investigators both during and after studies.
Sometimes it's the investigator's obligation to report two sets of findings: one that includes the data that was impacted by a protocol violation and another that does not include the data that was impacted by a protocol violation, Levin says.
If there is a large amount of noncompliance then that changes the study's results, and investigators will need to admit that this was a failed study and no conclusions can be drawn from it, he notes.
IRBs can discuss with investigators all of these scenarios when major protocol violations are discovered.
"In most cases where fraud or intentional malfeasance is found on the part of the investigator, then IRBs take the position that the study may not be published," Levin says.
Most often, IRB members will have more than one opinion on what should be done.
There isn't a black and white answer when an IRB discovers that a human subjects research site has collected some data that is tainted by a protocol violation.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