IRB director offers guidelines for running a more efficient office

Answer is data, data, data

The first step to improving your IRB office's efficiency is to collect data through a software system that will make sense of files and reports.

Next, make sure the reports include information about staffing levels, committee levels, and other workload issues, suggests Tanna MacReynold, CIP, institutional review office assistant director at the Fred Hutchinson Cancer Research Center in Seattle, WA.

"We've enhanced the database, and we've moved to an electronic review for IRBs," MacReynold says.

The electronic system, which the IRB office staff helped design, is called PIRO for Protocol Institutional Review Office, and it's helped considerably in making the office more efficient, MacReynold says.

"Before we had the new database the bulk of our work was expedited reviews, and we had to put together the expedited review agenda, which took one staff person three days to cut and paste the 40-some pages of expedited reviews," MacReynold says. "With the database change and the automated reports, it now takes about 30 seconds."

There also are efficiencies in the form of paperwork: for example, the office has reduced 26 hard copies of information down to one hard copy, and the database captures most of the information, allowing it to be viewed electronically through a Web-based system, MacReynold says.

"So it can be reviewed anywhere, so long as people have access to the Internet," MacReynold says. "They can go to a secure Web site and post real-time comments."

These comments from IRB members are tracked prior to the meeting, helping to inform members and facilitate discussions that will expedite the IRB's response, MacReynold adds.

IRB meetings are held monthly in the traditional face-to-face way, but the Internet commentary before the meetings is one of the electronic system's biggest benefits, she says.

Another major change that has facilitated more efficient work flow is an office reconfiguration. This is where data played an essential role.

Rather than just looking at the office's study output and basing staffing needs on that number, MacReynold looks at how many activities the IRB office performs over a 12-month period.

These include the following:

  • How many modifications and revisions are there?
  • How many annual reviews?
  • How many adverse events?
  • How many protocol violations, deviations, and other miscellaneous items?

MacReynold estimates necessary staffing levels based on the volume of these various office tasks over a five-year period.

"So I entered in the activities I wanted to check, and instead of just seeing that we had 1,200 protocols for 2005, I saw that we had 4,852 activities," MacReynold explains.

Then MacReynold could divide the activities by full-time equivalent (FTE) positions to come up with a formula of X number of activities and Y number of studies per FTE.

Since anecdotal evidence suggests the office's workload was manageable, but not distributed as efficiently as possible, MacReynold had data suggesting the number of activities per FTE that would be suitable. However, there still was room for improvement.

"It's not like I have a magic number that's right, but I can say that we switched things around, looked at different numbers, and then we came up with a new arrangement where we have an analyst and an assistant assigned to each committee," MacReynold says.

The analyst takes the IRB's meeting minutes, writes letters, and interacts with investigators, helping them make protocol improvements pre-review, she says.

The administrative assistant helps with processing paperwork after IRB approval and also does filing, and handles other tasks, MacReynold says.

As the operations person, MacReynold helps bring consistency to reviews between the center's several IRBs.

"If I'm sitting in on two or three committees, and one says, 'We should do it this way,' I might tell them that the other committee handled it a little differently," MacReynold explains. "So I bring consistency to the committees, and I bring institutional requirements and help with regulatory issues."

Another full-time staff person is the consent editor who helps investigators with their consents after they've written these, she adds.

An IRB also should help an institution decide when it's time to expand to additional IRB committees, based on the data about the workload.

In 2001, the center had about 700 studies, which were handled by two IRBs, MacReynold says.

When the number of studies expanded to 1,200, it was time to add a third IRB, she says.

"It was too much to ask each committee to review 600 protocols a piece," MacReynold says. "This was a different calculation; we already were growing our staff based on the activity level, but we weren't growing the committee."

Further calculations based on the data suggested it was a serious deficit to continue IRB activities with only two IRBs, she adds.

Once an IRB office has a handle on the data and can pull up the information that's necessary to tweak improvements and efficiencies, then it's a matter of managing the numbers and making adjustments, as needed, MacReynold says.

"Now that I have worked out these numbers, and we seem to be at a manageable workload with the numbers we have, then each year I will run the numbers, and the activities increase, I will compare this data to how manageable the workload feels to the staff," she explains.

If she were pressed to give a rule by thumb, then she would say that 157 protocols per FTE appear to be reasonable. Then you multiply the protocols by three to equal the average number activities, MacReynold says.

Another part of the information that MacReynold checks is the growth rate of the protocols versus the growth rate of activities.

"I want to see if activities are growing as fast as the protocols," she says. "If you just go by the protocols, you might put in for another FTE, but you won't know if it relates to your workload."

By following this comparison of growth rates, MacReynold found that the protocols were growing at a faster rate than the activities, so that provided a little more depth to her analysis of whether additional staff would be required.

Another factor is whether the protocols have greater proportions of high versus low risk, which could translate into more work versus less work, she says.

"Fifty protocols might mean something different between two institutions," MacReynold says.

"The other thing I did find was there is a whole lot of workload influences between the activities, and people have to consider these," she notes. "When you really get down to what the staff are doing, which is what I did, then there's all this hidden stuff that aren't accounted for."

For example, IRB office staff spend time on training IRB members, research staff, auditing, assisting subcommittees, and keeping up the Web site, and that time should be part of the equation, MacReynold says.

Nonetheless, the electronic system's implementation has resulted in staff work improvements.

Since the software system made the entire office more efficient, staff members have had time to do more substantive reviews and assessments, rather than spend their time attending to paperwork issues, she says.

"They can shift their knowledge and skills to looking at regulations, educating investigators, and those sorts of things," MacReynold says.

The software has become popular through word-by-mouth, although it is not sold or distributed publicly.

"At every single IRB conference, people ask if they can buy our software," MacReynold says. "It may be marketed in the future."