OMB seeks feedback on risk assessment guidelines

John Graham, head of the Office of Management and Budget's regulatory division, is making his last days memorable.

Graham, who leaves on March 1 to run the RAND Corp.'s graduate school in policy analysis, has issued a proposed risk assessment bulletin that would create new rules for regulators in the interest of promoting "sound science." An OMB spokesman said, "The quality and transparency of risk assessments currently varies from agency to agency. This bulletin provides clear, minimum standards for agency risk assessments."

The guidelines, which were published as a draft bulletin on Jan. 9, will be open to public comment through June 15 and have been forwarded to the National Academy of Sciences for review, according to an OMB statement. These inputs, as well as comments from federal agencies, will be used to issue final guidelines late this year.

The bulletin would apply to all federal agencies covered by the 1995 Paperwork Reduction Act, but is likely to have the most impact on the Consumer Product Safety Commission, Environmental Protection Agency, Nuclear Regulatory Commission, and the departments of Commerce, Health and Human Services, Interior, Labor and Transportation, according to an OMB official who asked to remain anonymous.

The bulletin would require agencies to explain in more detail the uncertainties involved in estimating hazards from things such as toxic chemicals and to present their findings ("this chemical can kill you") in the context of other risks ("you are more likely to be hit by a bus").

Environmentalists, who have portrayed Graham, a former Harvard professor, as a shill for industry, argue that burying risk statements under such caveats will make it more difficult to regulate anything.

But James Hammitt, director of the Harvard Center for Risk Analysis, said the proposed guidelines incorporate a good deal of common sense.

"I imagine that many of the best risk assessments that are done, for example, those at EPA that I'm most familiar with, already comply with most of this," Hammitt said. "These guidelines are trying to raise the level of the not-so-good examples."

The proposal covers a wide range of appraisals, including ones that evaluate baseline risk or risk mitigation activities, as well as less comprehensive studies such as exposure or hazard assessments. The bulletin separates risk assessments into two categories: "influential" and standard.

Assessments identified as influential--defined as having a "clear and substantial impact" on public policies or private sector decisions--would be subject to more rigorous standards.

The guidelines also would require in-depth risk analyses to include a range of estimates, including central and nonconservative estimates in addition to conservative "upper-bound" estimates. Assessments of the degree of uncertainty associated with each estimate would need to be included.

Hammitt said this could be difficult for studies of human exposure to rare toxic chemicals, which often are extrapolated from testing on laboratory animals, where direct applicability to humans is difficult to quantify.

A provision requiring agencies to consider all "significant comments" on draft risk assessments is likely to cause debate because it presumes that all "scientific" comments are significant, Hammitt said. "How do you decide if someone is a crackpot scientist?" he asked.