There were many changes between the Notice of Proposed Rulemaking (NPRM) and the final rule, and even more changes since the Advanced Notice of Proposed Rulemaking (ANPRM), which is a good thing, IRB experts say.

“The ANPRM was shoddy, and not well worked out,” says Erica Heath, CIP, a retired IRB director in San Anselmo, CA.

“The NPRM was better, but still pretty problematic; this version, however, is quite livable,” Heath adds.

The final rule also has an estimated annual reporting burden in table 21. It’s concise and readable, she says.

“This rule is literate, sensible, and meets their stated goals of reducing regulatory burden without really affecting protection, so it’s come a long way,” Heath says.

Another change is that the NPRM proposed to cover all biospecimens, regardless of their identifiability under the Common Rule, and the final rule does not require consent for secondary research with nonidentified biospecimens.

“They backed off from the worst in the NPRM,” Heath says. “The original two documents would have precluded a lot of very important research.”

For federally funded research using biospecimens, the rules are not as comprehensive as the original proposal, Smith says.

“The final rule is more concise about what you need to do, but leaves a lot to determine and more ways to segregate our federally funded studies from non-federally funded research,” says William Smith, JD, director of the IRB at Nova Southeastern University in Davie, FL.

“They wanted people to make more use of flexibility,” he adds.

Smith uses this example of the difference between the old Common Rule and the changes: “If you have a bucket of teeth, collected whenever someone had a tooth pulled, and you put these in a clinical depository,” he says. “There are no identifiers attached to it.”

If someone uses one of these teeth for research, then under the old regulations the teeth were not human subjects. But under the new Common Rule, if the tooth/biospecimen has an identifier and it’s used for certain purposes that involve federal funding, then there are more conditions attached to its use in research, Smith explains.

“So now it makes sense for institutions to distinguish between studies that use federal funding and those that don’t when they use data or specimens,” he says.

Before, research organizations had the option of checking or unchecking the box in how they treated non-federally funded research.

“Now, they’ve taken away the box and everyone’s box is unchecked,” Smith says. “That’s what the change in biospecimens will encourage institutions to do, creating a lot of requirements for biospecimens.”

It will result in creating, maintaining, updating, and giving access to registries and specimen banks that institutions will have to distinguish between federally funded and non-federally funded biobanks and repositories, he adds.

One of the most controversial — judging by its more than 300 comments — NPRM proposals was the one mandating that all institutions in the United States engage in cooperative research and rely on a single IRB as their reviewing IRB for that study.1

According to commentary in the final rule, the comments on this change divided research institutions, which were generally against the change, and scientific organizations, which were in favor of the change. The final rule adopts the NPRM proposal with modifications, but not the modification many IRBs and research institutions had desired — that the single IRB review is encouraged rather than mandated.1

Instead, the final rule allows flexibility in implementing the change, giving institutions the option of conducting additional internal IRB reviews for their own purposes. Those sorts of reviews would not have any regulatory status in terms of compliance with the Common Rule. The final rule also allows federal agency sponsors to select a central IRB, though lead institutions can propose the reviewing IRB and have their proposal approved by the sponsor.1

Further guidance will be needed, the final rule acknowledges.1

The net result will be confusion, especially since each central IRB arrangement will need to be ironed out legally, Smith predicts.

It can get complicated and frustrating because there is no template for such agreements, he adds.

“When you’ve got to worry about patient safety, who is going to do the local inspections and who has access to medical records?” Smith says. “Can you get the necessary documents, and will this streamline the review or will it be a case of every clinical trial having multiple negotiations over the contract, so it wouldn’t save time — just waste time?”

Research institutions and their legal offices will be wary of being held accountable for another IRB’s review, he adds.

“Lawyers are responsible for their clients’ best interest, and if a clinical trial goes south for whatever reason, you don’t want your client sued into bankruptcy because someone forgot to add a line in a consent form or to a clinical trial agreement,” Smith says. “That’s why they gave it three years to take effect.”

While cutting back on multiple IRB reviews is positive, the lack of a template for these reliance agreements is a major issue, and it will affect some IRBs, he adds.

“Most of the IRBs like ours — medium-sized — won’t close, but there will be a tendency to farm out reviews,” Smith says.

Independent IRBs, which are accustomed to the agreements, will benefit because they’re very good at handling these reviews quickly, he says.

“Small and medium-sized IRBs will have trouble keeping up the pace because they don’t have the staffing and expertise on hand,” he says.


  1. Federal Policy for the Protection of Human Subjects. Fed Reg. 2017-01058.