The trusted source for
healthcare information and
Assuring IRB submissions are mistake-free
Address policies, encryption, etc.
IRB application submissions often lack consistency, have omissions and errors, and other problems that IRBs should teach investigators to anticipate and correct before filing their application. Among these mistakes are policy errors, data security concerns, and other procedural problems.
An IRB expert offers this advice for how to correct and prevent the more common procedural and policy mistakes in IRB applications:
Unable to verify if research procedure described in protocol is consistent with university policies: "Sometimes we don't have enough information about the recruitment procedures to see if it's consistent with our policies," says Donna B. Konradi, PhD, RN, CNE, chair of the IRB at the University of Indianapolis and an associate professor in the School of Nursing at the University of Indianapolis in Indianapolis, IN.
Institutional policies regarding involving students in research, using campus labs, and providing protection in the event of participant injuries are all items that need to be addressed in the submission application when they're relevant to the described research projects.
For example, if a study involves using a campus physical therapy lab and a participant is injured during a procedure, then the way this adverse event is handled needs to be fully addressed.
"We need to know those procedures are built into the research protocol," Konradi says.
REMEDY: "We refer investigators to the policies on our website," she says.
Institutions often have policies on medical emergencies and accidents, student research participation, access to the research pool, and other issues that investigators need to check for consistency with what is in the IRB application.
Data security concerns: Investigators sometimes omit information or explanations about data security procedures they use, Konradi says.
"Sometimes their explanations are insufficient given the type of data they're collecting," she adds.
They need to describe their procedures for data storage and make certain their descriptions are consistent between documents.
For instance, the IRB submission should be clear on whether data are de-identified, coded, or anonymous.
If investigators are storing coded data, then the data should not be stored in the same file as the code because that's insufficient protection, Konradi says.
"We need to know if they have coded data, where they're storing the code, and what their security procedures are for the code, such as data encryption and password access," she adds.
REMEDY: "Our information services department in the last year has done a good job of educating the research community about our new encryption software," Konradi says. "This conversation is very important; you need to give people the heads up that security is a problem and something they need to think about."
Institutions can prevent data security issues with technological changes, as well.
For instance, new campus computers at the University of Indianapolis now have encryption software embedded, which minimizes data security risks.
"Over time, some of these problems will be greatly minimized from the faculty perspective," Konradi says. "We'll still have some student work and a problem with transporting data."
Qualifications of research team members: Researchers and staff need more than online CITI training, although that's a good starting place.
The IRB needs to know students involved in research projects have been appropriately trained to carry out their assigned activities.
"If they're administering a mini-mental status exam, we need to know they've been trained to do that," Konradi says. "There should be a brief paragraph describing the training."
Likewise, research staff members need to know how to operate all of the equipment that will be used in a research project, and the IRB should be told how they've been trained to handle cases of accidents and injuries, she adds.
REMEDY: IRBs should require investigators to post institutional procedures in labs and other locations where research procedures take place, Konradi suggests.
"They should train students and staff on the procedures and then post the procedure process in the lab so if there's a problem there is a ready reference available," she adds.
Investigators should provide staff qualifications on their IRB applications, showing that each person who will be performing a procedure has been trained or is otherwise qualified to do so.
Scientific design and design-related issues: Investigators sometimes submit applications that fail to make it clear how the study they propose to do is related to the state of the science or previous research, Konradi says.
"They haven't shown us how this study is the logical next step in the research and how it builds on what is identified in the literature," she explains. "You can't just go out and do something because you want to do it – there has to be a logical progression."
Sometimes IRB members can find no link or relationship between the research question and the types of data they are collecting, she notes.
"Investigators need to make clear connections between all those elements," she adds.
Another design issue involves investigators not differentiating between inclusion criteria and research-related interventions, Konradi says.
"Maybe the inclusion criteria would say that participants have to receive health clearance from their physicians," she says. "But the investigator asks participants to provide study-related data before obtaining medical clearance for participation from the physician."
REMEDY: Educate research staff and investigators about this issue.
Also, it helps for investigators to have someone proof-read their protocol and make sure they've connected all dots and links to the research question and proposed data collection.
"Sometimes it helps having a researcher who is not in your area of expertise look at the protocol and see if these elements are lined up in a way that's appropriate," Konradi says.
Research-ready documents: When an IRB submission involves an electronic survey, IRB members will want to see an active link to the survey. They'll want to be able to view the survey in the precise way that participants will view it, so just a print-out is not adequate for the review process.
Also, researchers might say their collection procedures are anonymous, but then they have the subject fill out forms in a way that is not anonymous, so the IRB needs to see the research-ready tools, Konradi explains.
"I want to see the survey as it fits into the electronic survey hosting system before doing the final IRB sign-off," she says. "And the informed consent documents need to be ready for an IRB stamp on the letterhead."
REMEDY: Again, educating researchers and staff is the main remedy.
Research institutions also might provide electronic letterheads so researchers could use these instead of photocopied or printed letterhead paper when writing informed consent documents, she notes.
IRBs can delay looking at links to surveys and other tools until the final IRB review of the research protocol, Konradi says.
"The committee might be okay to look at the survey on paper or in an electronic document form, but before the final sign-off, I need to see it in the hosting system," she adds.