Best Practices Spotlight
IRB improves workflow, eliminates hard stops
Redesign works for PIs
When an IRB redesigns its electronic submission system, it's the right time to consider redesigning the IRB's workflow as well, some experts say.
The key is to identify the hard stops in the IRB submission process and to design process improvements to eliminate them, says Michael Centola, MHS, CIP, IRB manager with the University of Massachusetts Medical School in Worcester. Centola was a co-author of a poster about IRB workflow, published at the 2013 Advancing Ethical Research Conference, held by the Public Responsibility in Medicine and Research (PRIM&R) Nov. 7-9, 2013, in Boston.
"The hard stops we were trying to avoid include, for example, having department chairs actively sign applications before they were submitted to the board," Centola says.
Previously, every application was emailed to the department chair, waiting for his or her signature before it could be submitted. This caused lengthy delays, especially when the email was overlooked or the department chair was on vacation, he explains.
With the workflow redesign, the new process is a passive approval approach: When an application is submitted to the IRB, the researcher's department chair receives an email stating that the researcher has submitted an application to the IRB and if the chair is OK with this application, he or she does not have to respond. If there is a question or problem, the chair may choose to call the IRB office immediately, Centola says.
"The idea behind this is that most department chairs have their own systems and a good awareness of what to expect when studies go to the IRB," he adds. "If they don't know about a study or already have similar studies, then they can contact the IRB."
Investigators were pleased with the change because it eliminated one of the frustrating delays in obtaining IRB approval.
"Their time to IRB submission was shortened," Centola says. "We're not collecting metrics, but some investigators say it was shortened by weeks."
But there also was some resistance to the change: Soon after the IRB switched to an electronic submission process and improved its workflow, some users complained that the new electronic system did not work intuitively, Centola says.
"The new system required some learning, and it was difficult for them to attend training sessions," he says.
The system would work easily for users once they had gone through the process once or twice because every step had the same workflow, he explains.
"Regardless of what they were doing — a request for exemption, a modification — they all had the same process," Centola says. "The only divergence was whether it went to expedited or full committee."
Once investigators used the system and grew accustomed to its logic, they were pleased with how it worked, he adds.
Here are some additional workflow improvements the IRB made:
• Passive approval for staff signatures: The IRB further improved workflow by using the passive approval approach for additional signatures required on IRB submissions. Previously, each person listed on the study staff needed to sign the application to participate in the research study, regardless of their role, Centola says.
Now, they can provide their approval to being listed by not responding: "If an investigator names me as the coordinator or research nurse, I will receive an email saying, 'Dr. Smith has added you to this application with the following role. Please log in and see if it's OK. If it's not OK, then you do nothing, but if you are not comfortable with what is listed, please contact the principal investigator immediately,'" he explains.
Previously, investigators had to bring the application all over the hospital to obtain signatures from as many as 20-30 people, Centola says.
"Now they don't have to spend time going around getting signatures," he adds.
The system appears to be working. Centola has been contacted by people who read the email and said they were unaware of the study. Then after they contacted the investigator at Centola's suggestion, they agreed to be listed, he says.
• Modifications to IRB applications: "Every time you add or remove someone from your IRB application, there is a modification sent to the IRB and they have to review and approve it," Centola says. "The modification to add someone to the study would be in our office for a couple of days."
The investigator would need to show that the person added to the study had completed all required training, and this might take additional time, he says.
The IRB eliminated this delay by checking with federal regulators and clarifying requirements, Centola says.
"They told us that if the roles of each staff member were listed in the body of the application, then taking an individual off the application and replacing that person with someone who matches that same role would not be a study modification," he explains.
"So now we have investigators describe the role, and if the staff person leaves and someone else comes in with the same certification and training, then it's not a modification to the research because that role was approved," Centola adds.
At continuing review, the IRB checks the study staff and applications to make sure all of the roles are filled appropriately.
"To date, we have not had any issues," Centola says. "Support staff and investigators find it such a huge benefit that they are not taking chances [with being out of compliance] by putting in people who have not done their training."
The IRB's workload also has greatly improved. Before, there would be 30-40 revisions or modifications a day to remove study staff, and now they can focus on more substantive work, he says.
"We have auditors who do routine audits, and there was only one situation where someone was added who did not do their CITI refresher course, and that person was working on it, so this change seems to be working," Centola adds.
• Eliminating duplication: The new electronic system eliminated duplication and redundancy. Once an investigator completed a form, he or she would not have to repeat that information in another form, Centola says.
The old process required the IRB to review a form, ask questions, and then have investigators complete another form to address all issues raised in the review, he explains.
"We'd return the form to them and it could take months," he says. "Now, we have one process for everything no matter what they're doing."
Investigators answer all required electronic fields, providing details, and the IRB can make decisions more quickly, and investigators have fewer forms to complete, he adds.
"We're now thinking about looking for more ways to improve the system," Centola says.