This afternoon, Congressman Brad Miller (D-NC), Chairman of the House Science Committee’s Subcommittee on Investigations and Oversight, will hold a hearing on recent revisions to the IRIS assessment process. IRIS (the Integrated Risk Information System) is EPA’s premier database of toxicological profiles for dangerous chemicals. The profiles are used for everything from setting cleanup standards at Superfund sites to determining liability in toxic tort suits. The problem is, IRIS only contains profiles for 548 chemicals. On average, 700 new chemicals enter commerce each year. Because IRIS numbers can serve as a cornerstone in the risk assessment/risk management process, an extensive database would greatly benefit policymakers in their daily work to protect public health.
We wanted to see whether some well-known toxins are adequately covered in the IRIS database, so we looked at the number of hazardous air pollutants that are listed in the Clean Air Act but missing from IRIS. We found that 17 percent of the chemicals have no profile at all. More remarkably, for a full two-thirds of the chemicals listed in the Clean Air Act, there is no data point in the IRIS database for the most relevant toxicological value needed to devise controls for toxic air pollution – the inhalation reference concentration (full report, press release)
EPA’s inability to populate the database with useful toxicological information for the Clean Air Act hazardous air pollutants (and many other environmental toxins) can be linked to problems with the administrative process that the agency has devised to develop IRIS profiles. To properly assess a chemical’s toxicity and develop an IRIS profile EPA scientists compile the best available scientific research, study and debate disparate and sometimes contradictory research findings, and consider the “weight of the evidence” to derive the numbers. All of this is done according to a step-by-step process, that has undergone numerous revisions in the last five years.
In the most recent revisions, EPA Administrator Lisa Jackson eliminated a number of procedures that the Bush Administration had devised to slow the process of updating the database. But, as our new report explains, she left in place one particular feature that has been a serious problem in the past – the “interagency review.”
Officially added to the IRIS assessment process in 2004, interagency review is billed as a way to ensure that federal agencies outside of EPA that have a particular interest in IRIS profiles can help ensure the scientific validity of the profiles. However, as we document in our new report and in a past report, (and as GAO has documented), interagency review is more of a hindrance than a help to IRIS staff. The problem is that the other agencies’ interest in IRIS profiles is usually tied to a fear of future regulation or cleanup costs, so their influence over the assessment process is typically driven by ulterior motives like delaying regulation and minimizing risk characterizations. In other words, other federal agencies behave – and should be treated – more like private-sector stakeholders. They should not be given privileged access to EPA staff or officials. They should provide their comments during the standard public comment period. Each IRIS assessment goes through EPA’s intra-agency review, plus an external review by scientific experts. Additional layers of review for outside federal agencies are not necessary. As we conclude in the new report, the entire interagency review process undermines the scientific integrity of the IRIS process, and EPA should abandon it.