By Melinda Young, Author
Bioethicists and the human research protections industry have been shocked in recent years by reports of rogue researchers carrying out investigations that raise huge red flags and, in some instances, could be regulatory breaches.
In 2018, the scientific world learned that a Chinese investigator claimed his research resulted in genetically edited newborn twins. He also claimed his research was approved by a Chinese ethics board. (For more information, see story in the January 2019 issue of IRB Advisor.)
In 2017, William Halford, PhD, a professor at Southern Illinois University School of Medicine, had conducted herpes clinical trials on the island of St. Kitts. Halford died that year from a rare cancer, leaving behind his research and its controversial way of circumventing regulatory safeguards. He had not gotten his research approved by the university’s IRB, according to university officials interviewed by IRB Advisor in 2017 but who have not responded to follow-up interview requests. (See story about Halford’s research in the November 2017 issue of IRB Advisor.)
While these are two highly publicized examples of potential ethical and regulatory breaches, they are part of a trend, says Joseph Crossno, MS, CPIA, assistant director in the office of research compliance at Central Michigan University in Mt. Pleasant.
“In 2015, we started getting anecdotal reports and seeing stories in the press regarding research — either approved or unethical research — being conducted in various places around the country,” he says.
Crossno recalls seeing social media reports from researchers about studies they had started at another university but continued with after they left, possibly without informing their IRBs.
“It had us asking the question, ‘Could this happen here?’ And if it did, what was the likelihood we would know in time to do something about it?” he says.
These concerns prompted Crossno to help lead an effort to develop standard operating procedures that could help prevent rogue research.
“We started monitoring social media outlets, including Facebook, Twitter, LinkedIn,” he says. “Some people were advertising things on other platforms for a time, but Facebook was the popular one with human subject research for recruitment and advertising.”
The goal behind the monitoring was to check on what the IRB had approved for a study and then see whether the investigator was recruiting or advertising the study as had been approved by the IRB, says Robert S. Bienkowski, PhD, CIP, CHRC, director of the office of research compliance at Central Michigan University. (See story on IRBs and recruitment ads in this issue.)
“It’s reasonable to ask the question, ‘How do you know unapproved research is not going on at your university?’” Bienkowski says. “A careless administrator might say, ‘You just don’t know, and we haven’t seen any pictures or stories in the media, so we just don’t know.’”
By starting a process of surveillance and monitoring, a research institution can discover potential breaches and problems, he adds.
“In classic compliance processes, you ask about the risks to the institution, including the risk of the institution’s reputation, and then you develop a process to address the risk,” Bienkowski explains.
The Central Michigan University research compliance office conducts surveillance, looking for unknown problems. “We don’t know what it is, but we know it’s out there,” Bienkowski says.
The program found that most noncompliance issues involved unapproved changes in recruitment materials, Crossno says.
Surveillance is a proactive way to monitor research compliance, including recruitment advertising and marketing.
“We designed the surveillance as a very simple, not labor-intensive program,” Crossno says. “We made it as simple as possible to start and planned to tweak things and ramp up the program as we go along.”
For instance, the surveillance program uses automatic electronic search alerts, such as Google Alerts.
“We put in half a dozen search terms and get automatic reminders,” Crossno says. “To start, I put in ‘research’ and ‘Central Michigan University,’ and that generated a lot of hits. I’d spend all day looking through the returns.”
Now, Crossno uses keywords that are more refined, based on the results they saw in the first few years of the surveillance program. This resulted in findings of mostly minor noncompliance, he adds.
For example, surveillance sometimes finds problems with student-initiated studies in recruitment and advertising. Often, these are cases of student-initiated research in which the advertising either wasn’t approved by the IRB, or it was approved and later modified in a way that removed key elements, Crossno explains.
“All of these findings have been minor noncompliance,” he says.
Sometimes, researchers will talk about their research plans on a blog, outlining their plans for the months ahead. If their blog description doesn’t match what they submitted in a proposal to the IRB, Crossno will reach out to the faculty researcher, student researcher, and faculty adviser and say, “You’ll need to modify your protocol if you’re really going to do this,” he says.
“I try, almost exclusively, to go through the faculty member,” Crossno says. “Sometimes, we’ll contact a student but copy the faculty member.”
If Crossno knows the student investigator, he might include the student on the email message, writing, “You posted this flier on this platform, and I just wanted to make sure you and Dr. So-and-So are aware this is not the approved version.”
When investigators take actions in their studies that an IRB has not approved, it is protocol drift, Bienkowski says.
Researchers sometimes do not understand that the IRB has to see everything that will be shown to participants, including changes to the size of font, color, layout of recruitment materials, and so forth, he explains.
These instances of noncompliance found through surveillance often involve minor changes to what investigators presented to the ethics board. From the researcher’s perspective, these changes seem trivial: “There is a factor of ‘Do I really have to go back to the IRB to say I reformatted the original application, and it looks better now?’” Bienkowski says. “But yes, they do.”
When noncompliance occurs, whether minor or serious, the situation is brought to the attention of the oversight committee, including the IRB if it is human subjects research, Bienkowski says. “The chair of the committee will make an initial determination,” he adds.
One example of a noncompliant study advertisement involves a case in which the investigator placed a banner in front of the student union. The banner advertised participation in the survey and emphasized participant compensation, the dates of the study, and how participants would be contacted and what they had to do, Crossno says.
“The problem was the study wasn’t approved yet, and the emphasis was on compensation. The principal investigator’s name and location were missing,” he explains. “That was an example of one that was definitely noncompliant, and we referred it to the IRB chair.”
Since it was minor noncompliance and there was no impact on subject risks, it could be handled by the IRB chair, Crossno notes.
“Because advertisements had become such a recurring issue, the chair asked me to give a talk at the convened IRB about advertisements and what had to be included and excluded,” Crossno says. “So when the IRB reviews advertising and recruitment materials, this gives them a checklist and reminder with quick bullet points of what they need to look for.”
Compliance monitoring involves collecting data, but this does not have to be an exhaustive process. “One thing we did to keep our version of this program simple is I don’t invest a significant amount of time tracking statistics,” Crossno says. “I take snapshots over brief periods of time.”
The program also relies on communication and retraining/education of researchers and their faculty advisers.
“I sit down with a faculty member or student and say, ‘Here’s what we found. Here’s what’s required. Here’s what you did, and in the future, you need to make sure you do what’s required,’” Crossno says. “These are not hostile sessions; we just assess the researcher’s level of understanding, and they learn what is needed going forward.”