In a recent exchange with a colleague, he commented on what he called "Guidance Creep". He was referring to the situation where OHRP (or some other regulatory agency) responds to a question about an IRB practice and it is disseminated as "OHRP says..." which then becomes interpreted as OHRP guidance.
"Guidance Creep" is actually worse than "Regulatory Creep" since it is harder to refute. For regulatory creep, if it's not in the regs, it's not in the regs. But, with guidance creep, once something is seen as "best practice" it becomes what is expected in the field. Institutions have to remember that the only guidance that is "binding" is the official guidance that posted on the OHRP website. Unofficial opinions are not to be interpreted as guidance. Even with official guidance, it is still guidance. If an institution can make a good case for why deviating from the guidance adequately complies with the regulations and protects subjects, then they are free to do so. Of course, OHRP may not agree with the justification, but it rarely "second guesses" a reasoned justification.
This is also a problem with the accreditation process. I spend a lot of time with clients who are going through accreditation trying to distinguish between what's required in the regulations and what AAHRPP expects. When it is an AAHRPP expectation, the institution can decide if it works for them and, if it doesn't, push back. Of course, there is a risk to that if AAHRPP doesn't accept your argument, so the institution has to decide if pushing back is worth the risk. That, though, is a good thing since it requires that the institution really think through its practices and procedures and decide what's really important for human research protections.
All-in-all, it is important for institutions to seriously consider their policies and practices and not just blindly "comply with the regulations" or adopt OHRP or AAHRPP "guidance".