Monday, October 27, 2008

Report from SACHRP, October 2008

Today I attended the public meeting of the Secretary's Advisory Committee on Human Research Protections (SACHRP). Most of the day's discussion concerned the slow pace with which SACHRP's recommendations over the years have been implemented in any way; some have been sitting for years without action. OHRP officials offered detailed explanations of the complexity of policy-making process and OHRP's lack of resources. While these issues help explain OHRP's neglect of questions important to the social sciences and humanities, today's discussion had little of direct importance to scholars in those fields.

Here are a few tidbits of potential significance.

What is the difference between guidance and regulation?

Christian Mahler, a lawyer with HHS's Office of General Counsel, explained that only regulations can be enforced; as a legal matter, institutions are free to ignore OHRP guidance documents. But he and other officials acknowledged that in practice, the difference may not be great. As acting OHRP director Ivor Pritchard conceded, "when we issue guidance . . . people look at every phrase, clause, use of punctuation to see what was meant by OHRP.” He noted that institutional officials may believe the safest course is to comply with all OHRP guidance.

Pritchard said that OHRP tries to differentiate, when drafting guidance, between "must" statements (that indicate OHRP's interpretations of the regulations) and "should" statements that can be ignored if an institution has good reason. He also stated that the best course might be to issue guidance documents that offer multiple ways to comply with the regulations, though it's not clear that OHRP has ever issued such guidance.

Finally, Mahler pointed to the Office of Management and Budget's 2007 "Final Bulletin for Agency Good Guidance Practices.” That bulletin notes that

The courts, Congress, and other authorities have emphasized that rules which do not merely interpret existing law or announce tentative policy positions but which establish new policy positions that the agency treats as binding must comply with the [Admistrative Procedure Act]’s notice-and-comment requirements, regardless of how they initially are labeled. More general concerns also have been raised that agency guidance practices should be better informed and more transparent, fair and accountable. Poorly designed or misused guidance documents can impose significant costs or limit the freedom of the public.

Despite that last sentence, the bulletin seems designed more to limit regulation of economic affairs than to safeguard the freedom of the public. Still, if its approach were followed, OHRP might not be able to get away with so many arbitrary decisions.

What is research?

In February 2007, then-OHRP director Bernard Schwetz told the New York Times that OHRP would, by the end of 2007, issue guidance on what is and is not research under the regulations. OHRP is almost a year overdue in keeping that promise, but apparently it is still at work. Pritchard noted that "We have been working on a guidance document on the definition of ‘research’ for several years now.” Pritchard did not indicate how far along that process is, or whether OHRP will solicit public comment before promulgating that document.

What is an ideal consent process?

In her presentation, Elizabeth Bankert, co-chair of SACHRP's Subpart A Subcommittee, addressed the problem of IRB insistence on long, complex consent forms. She noted that while the regulations requiring consent forms have not changed since 1991, IRBs have been demanding ever more detailed forms, and have gained a reputation for "wordsmithing" and "nit-picking." Not only does this erode investigator trust and respect for IRBs, but it also "diminishe[s] the consent process for subjects."

Bankert--drawing on work by subcommittee member Gary Chadwick--challenged the presumption that "the form must contain every piece of information and in the same detail as required in the consent process," which leads to such long forms. She called upon OHRP and FDA to endorse the use of shortened consent forms: about 3-4 pages for a clinical trial, and just one for "surveys, etc."

Bankert offered these suggestions as a way to decrease some burdens on IRBs and investigators, but she then suggested that IRBs would still need to review the consent process. She did not explain how the same incentives for nit-picking wouldn't lead to endless fretting about the consent process, rather than the consent form. As SACHRP member Jeffrey Botkin pointed out, IRBs and investigators turn to long forms out of self-preservation. And as David Foster noted, FDA inspectors and AAHRPP accreditors have demanded ever longer forms. Before Bankert and Chadwick can address the problem of silly forms, they need to understand the system that produces them.

Predictably, almost all the discussion of consent forms centered around clinical trials. For example, Bankert offered sample executive summaries for a drug trial and a request to store blood or tissue for future research, but no comparable document for a social science project. And one committee member, expressing her concerns, forgot to speak of "human subjects" and instead started talking about protecting "patients." Given the complexity of the issue, and the dominance of SACRHP by medical researchers and bioethicists, I would expect that any change will be designed for clinical trials, and then imposed on social researchers.

No comments: