Friday, September 23, 2011

Would ANPRM's Audits Require Massive Record-Keeping?

Part of the ANPRM's proposal for a new category of Excused research is an audit regime for projects that researchers declare eligible for Excusal. An observer asks whether this requirement would impose inappropriate burdens on researchers. I am hopeful that it would not, but commenters on the ANPRM should be aware of this possibility as they answer question 21.

The ANPRM explains:

We are considering a mechanism to track Excused research, and to audit only a small but appropriate portion of such research, because it would still be subject to other regulatory protections, such as the proposed data security and information protection standards and certain consent requirements. In addition, such a mechanism to track and audit Excused research will also enable institutions to assure that the research does indeed meet the criteria for inclusion in the Excused category. (That is all that an audit would in most cases involve: a brief review of the registration form, similar to what many institutions currently do when they determine whether a study is exempt.) Key to this would be a requirement that researchers register their study with an institutional office by completing a brief form. This would make the institution aware of the research and identify the study’s principal investigator. In addition the institution could choose to review some of the submissions at the time they are filed (and we contemplate that this would only be done in a relatively small percentage of the filings) and if deemed appropriate, require that the study be sent for expedited review or, in exceptionally rare cases, convened IRB review.

The proposed auditing requirement is intended to encourage institutions to use the regulatory flexibility proposed for the Excused category of research. Rather than maintaining many institutions’ current practice of routinely requiring that research that meets the current exemption categories undergo some type of review before it is permitted to proceed, the proposed auditing requirement would provide institutions with information needed to assess their compliance with the new Excused category without unnecessarily subjecting all such research to either prospective review, or even routine review sometime after the study is begun.

Writing on the IRB Forum, Dr. John H. Noble Jr. warns that such an audit mechnanism could impose new burdens:
Maybe I can offer some insight about conducting audits from efforts of the government auditors I worked with in the 1970s-1980s. Far too often the auditors conclude that an audit cannot be completed because of lack of needed records. Ironically, the proposed new Common Rule requiring random retrospective audits of exempt studies, unless accompanied by very specific record-keeping requirements for the various kinds of research that might be considered "exempt," as well as the noted institutional uncertainty and concern about "regulations, laws, obligations and risks," seems likely to add to the burden of PIs, the research institutions, sponsors, and IRBs. Will the lack of a specific document or record create an "audit exception" that is deemed blameworthy? That said, in my experience the common reason that organizations were able to walk away from a government audit as "not susceptible to audit for lack of needed records" was to avoid disclosure of illegalities. What happens when the prosecution lacks the evidence to prosecute? Check out what penalties can be imposed for lack of adequate record-keeping. For a start, see, for example: and and,,id=183208,00.html and .

Yummy, I see adoption of such a rule as guaranteeing full-employment for lawyers and auditors who will flock to help the thousands of research institutions that exempt certain studies as SOP. I would also add statisticians, who will have to design adequate size samples of the exempt studies in order to satisfy the audit requirements in this regard. Some of us government retirees with program evaluation and audit experience may well be drawn out of retirement to fill the demand for expertise to implement the proposed new Common Rule--if it is serious about moving in this direction. It is important to keep in mind that when audits and auditors become part of regulatory equation, it is only a matter of time before government audit standards will be evoked or modified to encompass the new activity. Be careful of what you pray for!!!

I am hopeful that the human-research audit regime need not be nearly as elaborate as the financial and health analogies Dr. Noble offers. It could instead resemble the retrospective audits now used in cases of alleged plagiarism, fabrication, and falsification. Researchers are not subject to "very specific record-keeping requirements" beyond general scholarly norms that they retain enough notes and other materials that they can answer probing questions about their research. This system is not perfect (consider the debate over Michael Bellesiles's soggy notepads), but I think it suffices.

Moreover, I would think that in most cases, an audit could be conducted using solely published materials. That is, if you declare that your research is Excused because you are only talking to people, and then you publish an article about the LSD you gave them, that would be a problem.

Still, given the history of IRB regulations so far, it would be a mistake to ignore Dr. Noble's nightmare. I recommend that replies to question 21 suggest that the auditing system not impose any new record-keeping requirements. And I agree that the proposal for "random selection" is odd. Why not limit audits to cases in which institutions have reason to suspect noncompliance as is now the case with plagiarism, fabrication, and falsification?

[Note: In deference to the IRB Forum's policy I secured Dr. Noble's permission to quote his posting.]

No comments: