On October 6, 2011, the House Energy and Commerce Subcommittee on Environment and the Economy held a hearing on 'Chemical Risk Assessment: What Works for Jobs and the Economy?' According to the background memorandum, the hearing was intended to explore the U.S. Environmental Protection Agency's (EPA) Integrated Risk Information System (IRIS), and the relationship between IRIS and EPA's regulatory requirements, 'specifically whether IRIS is producing high quality science-based risk assessments that are suitable for regulatory objectives, or policy judgments that could harm businesses and the public.' IRIS has been criticized for years, but has come under intense criticism and Congressional scrutiny lately. The National Academy of Sciences (NAS) released a report in April 2011 on EPA's IRIS assessment for formaldehyde. Although EPA had asked NAS to review only the IRIS assessment, NAS included general comments on EPA's IRIS development process. On July 14, 2011, the House Science, Space, and Technology Subcommittee on Investigations and Oversight, prompted in part by NAS's report, held a hearing to better understand the process behind the development of IRIS assessments, whether EPA plans on adopting NAS's recommendations, and whether EPA assessments are based on the best available evidence and evaluated in accordance with established protocols. Immediately prior to the hearing, on July 12, 2011, EPA announced plans to improve the IRIS process. The background memorandum for the October 6, 2011, hearing and witness testimony are available online.
Witnesses on Panel 1 included:
- Dr. Paul Anastas, Assistant Administrator for the Office of Research and Development (ORD), EPA;
- Mr. David C. Trimble, Director, Natural Resources and Environment, U.S. Government Accountability Office (GAO); and
- Dr. David C. Dorman, Dean for Research and Graduate Studies, North Carolina State University, on Behalf of NAS.
Witnesses on Panel 2 included:
- Dr. Michael Honeycutt, Director of Toxicology Division, Texas Council on Environmental Quality (TCEQ);
- Harvey Clewell, Ph.D., DABT, The Hamner Institutes for Health Science;
- Mr. Jerry A. Cook, Technical Director, Chemical Products Corporation (CPC); and
- Dr. Thomas A. Burke, Associate Dean for Public Health Practice and Training, Department of Health Policy and Management, The Johns Hopkins Bloomberg School of Public Health.
Anastas summarized NAS's overall recommendations and EPA's responses, but added little new information to ORD's defense of the IRIS Program. According to Anastas, EPA is working closely with its Science Advisory Board (SAB) 'on how to bring to bear its expertise on an ongoing basis to focus on the quality, transparency and scientific rigor of IRIS assessments and guide EPA's response to the NAS recommendations.' Anastas noted that, while EPA's overarching goal is to improve IRIS assessments, the improvements will have a greater impact on new assessments rather than those already in development. Anastas stated: 'It is important to note that the NAS report viewed the implementation of their recommendations as a multi-year process.'
Trimble discussed GAO's past work on the timeliness and credibility of IRIS assessments and EPA's May 2009 IRIS assessment process. GAO is currently reviewing the 2009 assessment process and EPA's progress in implementing it and intends to issue a report later in 2011. GAO identified the following factors that it believes collectively could present significant management challenges to EPA's ability to complete timely, credible IRIS assessments:
- Unlike other EPA programs with statutory deadlines for completing activities, no enforceable deadlines apply to the IRIS Program. GAO believes the absence of statutory deadlines may contribute to EPA's failure to complete timely IRIS assessments;
- Because science and methodologies are constantly changing, there will always be a tension between assessing the best available science and waiting for more information. The IRIS Program will remain viable only if it continues to use the best science available at the time of its assessments and plans for periodic updates of assessments to identify the need for revisions;
- An overarching factor that affects EPA's ability to complete IRIS assessments in a timely manner is the compounding factor of delays;
- Long-standing difficulties in completing assessments of chemicals of key concern stem in part from challenges by external parties, including those that may be affected by EPA regulation of chemicals should an assessment lead to such action. Such challenges are to be expected and can be best addressed by EPA's focusing on the best available science, obtaining credible expert review, and completing the assessments; and
- IRIS process reforms are not established in regulation or statute and thus can easily be altered. Continual changes to the process have presented a challenge to the chemical managers who undertake the assessments. To produce timely, credible IRIS assessments over a sustained period of time, it will be important for EPA to maintain a stable, consistent process going forward.
Dorman testified that NAS, in preparing its report on EPA's formaldehyde assessment, found recurring methodologic programs that cut across the components of its charge. NAS thus commented on EPA's general methodology and offered general suggestions with regard to the processes EPA uses to develop IRIS assessments. Dorman noted that the general problems that NAS identified are not unique and have been reported over the years by other NAS National Research Council (NRC) committees charged with reviewing other IRIS assessments. According to Dorman, problems with clarity and transparency of the methods appear to be a repeating theme. Dorman stated that although EPA announced plans to revise the overall IRIS process, the focus of the revisions appears to be on the steps taken after an assessment has been generated (the multiple layers of review), and the NAS's focus was on the completion of the draft IRIS assessment (the development phase).
According to Honeycutt, because of the lack of scientific defensibility and the implications of EPA's new chemical assessments, TCEQ decided to develop its own chemical assessments. TCEQ chose this approach after EPA's assessments became 'more precautionary in nature in lieu of relying on scientific data.' Honeycutt cited EPA's IRIS assessment for formaldehyde, under which the formaldehyde in a person's breath would be more than five times higher than the highest level that EPA would call safe. He also noted that typical soil and water concentrations of chemicals, including those naturally occurring, would be considered unsafe, which would have significant implications for remediation programs. Honeycutt submitted as part of his testimony copies of technical comments that TCEQ has submitted to EPA outlining the 'numerous scientific shortcomings of recent EPA chemical risk assessments.'
Clewell described his concern that the lack of objectivity and transparency in some recent IRIS assessments would impair the ability of decision makers to make informed risk management decisions. He noted that some recent assessments, such as those for inorganic arsenic, formaldehyde, and dioxin, present only a low-dose linear default approach that assumes these chemicals are carcinogenic at any concentration, despite strong evidence that the true dose-response is nonlinear and that the default greatly overestimates the actual risk at current human exposure levels. Clewell testified that decision makers would be better informed by a balanced and objective discussion of both alternatives and the presentation of analyses based on both alternative approaches in the risk characterization section of the assessment. Clewell stated that, in a number of cases, NAS and SAB peer reviews have requested EPA modify its IRIS assessment to present objectively multiple risk assessment options, but EPA has not complied. According to Clewell, presenting only a conservative default approach, when there is a viable alternative, provides the decision maker 'with an inaccurate characterization of risk that compromises his ability to make informed risk management decisions.
Cook testified regarding the decline in demand for CPC's barium products. CPC attributes the decline to the over-regulation of barium by EPA under the Resource Conservation and Recovery Act (RCRA). Clewell stated that although IRIS should function as an up-to-date repository of chemical toxicology assessments to be relied upon as the scientific basis for regulatory decision making throughout EPA, CPC found IRIS to be unresponsive to new data showing that barium was significantly less toxic than previously assumed. According to Clewell, it took 11 years and actions resulting from submission of Requests for Correction and Reconsideration under the Information Quality Act to achieve an IRIS assessment that reflected sound science. In July 2005, EPA finally revised the IRIS assessment to recognize the 1994 National Toxicology Program study as the principal study defining chronic barium toxicology and identifying a higher no observed adverse level.
Burke addressed three points in his testimony:
- Risk assessment is an essential tool that informs public health policy decisions far beyond the regulatory mandates of EPA;
- The EPA IRIS Program plays a critical role in providing the scientific assessment of hazards, yet there is an important distinction between the IRIS assessment process and the ultimate risk management decision; and
- The NAS 2009 report Science and Decisions provides a framework for improving the risk assessment process and addressing the needs of decision makers.
The scheduling of the hearing reflects continued Congressional concern with the quality of IRIS assessments, the lack of transparency with their development, and the time ORD takes to prepare each. On the whole, Anastas provided no new information on EPA's claimed commitment to address the concerns that have been identified by NAS and others, nor did Anastas seem especially receptive to the notion these problems exist and need immediate remediation. That said, several of the Subcommittee members' questions elicited spirited debate on the relevance of IRIS assessments, whether they are 'risk' or 'hazard' assessments (ORD vigorously believes, despite the IRIS name, that they are hazard assessments), their utility to risk managers, and how exactly ORD is addressing the NAS recommendations to improve the quality of IRIS assessments.
That said, the hearing properly focused much needed attention on deficits in the IRIS Program. The hope is continued Congressional scrutiny will hasten the fundamental changes to the IRIS Program that are long overdue.