Chemical Assessments
EPA's New Assessment Process Will Further Limit the Productivity and Credibility of Its Integrated Risk Information System Gao ID: GAO-08-810T May 21, 2008The Environmental Protection Agency's (EPA) Integrated Risk Information System (IRIS) contains EPA's scientific position on the potential human health effects of exposure to more than 540 chemicals. Toxicity assessments in the IRIS database constitute the first two critical steps of the risk assessment process, which in turn, provides the foundation for risk management decisions. Thus, IRIS is a critical component of EPA's capacity to support scientifically sound environmental decisions, policies, and regulations. This testimony discusses (1) highlights of GAO's March 2008 report, Chemical Assessments: Low Productivity and New Interagency Review Process Limit the Usefulness and Credibility of EPA's Integrated Risk Information System, and (2) key aspects of EPA's revised IRIS assessment process, released on April 10, 2008. For the March 2008 report, GAO reviewed and analyzed EPA data and interviewed officials at relevant agencies, including the Office of Management and Budget (OMB). For this testimony, GAO supplemented the prior audit work with a review of EPA's revised IRIS assessment process announced on April 10, 2008.
In its March 2008 report, GAO concluded that the IRIS database is at serious risk of becoming obsolete because EPA has not been able to routinely complete timely, credible assessments or decrease its backlog of 70 ongoing assessments--a total of 4 were completed in fiscal years 2006 and 2007. In addition, recent assessment process changes, as well as other changes EPA was considering at the time of GAO's review, further reduce the timeliness and credibility of IRIS assessments. EPA's efforts to finalize assessments have been thwarted by a combination of factors, including two new OMB-required reviews of IRIS assessments by OMB and other federal agencies; EPA management decisions, such as delaying some assessments to await new research; and the compounding effect of delays-even one delay can have a domino effect, requiring the process to essentially be repeated. The two new OMB/interagency reviews of draft assessments involve other federal agencies in EPA's IRIS assessment process in a manner that limits the credibility of IRIS assessments and hinders EPA's ability to manage them. For example, the OMB/interagency reviews lack transparency, and OMB required EPA to terminate five assessments EPA had initiated to help it implement the Clean Air Act. The changes to the IRIS assessment process that EPA was considering, but had not yet issued at the time of our review, would have added to the already unacceptable level of delays in completing IRIS assessments and further limited the credibility of the assessments. On April 10, 2008, EPA issued its revised IRIS assessment process, effective immediately. In its February 2008 comments on GAO's draft report, EPA said it would consider the report's recommendations, which were aimed at streamlining the process and better ensuring that EPA has the ability to develop transparent, credible assessments. However, EPA's new process is largely the same as the draft GAO evaluated, and some key changes are likely to further exacerbate the productivity and credibility concerns GAO identified. For example, while the draft process would have made comments from other federal agencies on IRIS assessments part of the public record, EPA's new process expressly defines such comments as "deliberative" and excludes them from the public record. GAO continues to believe that it is critical that input from all parties--particularly agencies that may be affected by the outcome of IRIS assessments--be publicly available. In addition, the estimated time frames under the new process, especially for chemicals of key concern, will likely perpetuate the cycle of delays to which the majority of ongoing assessments have been subject. Instead of significantly streamlining the process, which GAO recommended, EPA has institutionalized a process that from the outset is estimated to take 6 to 8 years to complete. This is problematic because of the substantial rework such cases often require to take into account changing science and methodologies. Since EPA's new process is not responsive to GAO's recommendations, the viability of this critical database has been further jeopardized.
GAO-08-810T, Chemical Assessments: EPA's New Assessment Process Will Further Limit the Productivity and Credibility of Its Integrated Risk Information System
This is the accessible text file for GAO report number GAO-08-810T
entitled 'Chemical Assessments: EPA's New Assessment Process Will
Further Limit the Productivity and Credibility of Its Integrated Risk
Information System' which was released on May 22, 2008.
This text file was formatted by the U.S. Government Accountability
Office (GAO) to be accessible to users with visual impairments, as part
of a longer term project to improve GAO products' accessibility. Every
attempt has been made to maintain the structural and data integrity of
the original printed product. Accessibility features, such as text
descriptions of tables, consecutively numbered footnotes placed at the
end of the file, and the text of agency comment letters, are provided
but may not exactly duplicate the presentation or format of the printed
version. The portable document format (PDF) file is an exact electronic
replica of the printed version. We welcome your feedback. Please E-mail
your comments regarding the contents or accessibility features of this
document to Webmaster@gao.gov.
This is a work of the U.S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed
in its entirety without further permission from GAO. Because this work
may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this
material separately.
Testimony before the Subcommittee on Investigations and Oversight,
Committee on Science and Technology, House of Representatives:
United States Government Accountability Office:
GAO:
For Release on Delivery:
Expected at 11:00 a.m. EDT:
Wednesday, May 21, 2008:
Chemical Assessments:
EPA's New Assessment Process Will Further Limit the Productivity and
Credibility of Its Integrated Risk Information System:
Statement of John B. Stephenson, Director:
Natural Resources and Environment:
GAO-08-810T:
GAO Highlights:
Highlights of GAO-08-810T, testimony before the Subcommittee on
Investigations and Oversight, Committee on Science and Technology,
House of Representatives.
Why GAO Did This Study:
The Environmental Protection Agency‘s (EPA) Integrated Risk Information
System (IRIS) contains EPA‘s scientific position on the potential human
health effects of exposure to more than 540 chemicals. Toxicity
assessments in the IRIS database constitute the first two critical
steps of the risk assessment process, which in turn, provides the
foundation for risk management decisions. Thus, IRIS is a critical
component of EPA‘s capacity to support scientifically sound
environmental decisions, policies, and regulations.
This testimony discusses (1) highlights of GAO‘s March 2008 report,
Chemical Assessments: Low Productivity and New Interagency Review
Process Limit the Usefulness and Credibility of EPA‘s Integrated Risk
Information System, and (2) key aspects of EPA‘s revised IRIS
assessment process, released on April 10, 2008. For the March 2008
report, GAO reviewed and analyzed EPA data and interviewed officials at
relevant agencies, including the Office of Management and Budget (OMB).
For this testimony, GAO supplemented the prior audit work with a review
of EPA‘s revised IRIS assessment process announced on April 10, 2008.
What GAO Found:
In its March 2008 report, GAO concluded that the IRIS database is at
serious risk of becoming obsolete because EPA has not been able to
routinely complete timely, credible assessments or decrease its backlog
of 70 ongoing assessments”a total of 4 were completed in fiscal years
2006 and 2007. In addition, recent assessment process changes, as well
as other changes EPA was considering at the time of GAO‘s review,
further reduce the timeliness and credibility of IRIS assessments.
* EPA‘s efforts to finalize assessments have been thwarted by a
combination of factors, including two new OMB-required reviews of IRIS
assessments by OMB and other federal agencies; EPA management
decisions, such as delaying some assessments to await new research; and
the compounding effect of delays–even one delay can have a domino
effect, requiring the process to essentially be repeated.
* The two new OMB/interagency reviews of draft assessments involve
other federal agencies in EPA‘s IRIS assessment process in a manner
that limits the credibility of IRIS assessments and hinders EPA‘s
ability to manage them. For example, the OMB/interagency reviews lack
transparency, and OMB required EPA to terminate five assessments EPA
had initiated to help it implement the Clean Air Act.
* The changes to the IRIS assessment process that EPA was considering,
but had not yet issued at the time of our review, would have added to
the already unacceptable level of delays in completing IRIS assessments
and further limited the credibility of the assessments.
On April 10, 2008, EPA issued its revised IRIS assessment process,
effective immediately. In its February 2008 comments on GAO‘s draft
report, EPA said it would consider the report‘s recommendations, which
were aimed at streamlining the process and better ensuring that EPA has
the ability to develop transparent, credible assessments. However,
EPA‘s new process is largely the same as the draft GAO evaluated, and
some key changes are likely to further exacerbate the productivity and
credibility concerns GAO identified. For example, while the draft
process would have made comments from other federal agencies on IRIS
assessments part of the public record, EPA‘s new process expressly
defines such comments as ’deliberative“ and excludes them from the
public record. GAO continues to believe that it is critical that input
from all parties”particularly agencies that may be affected by the
outcome of IRIS assessments”be publicly available. In addition, the
estimated time frames under the new process, especially for chemicals
of key concern, will likely perpetuate the cycle of delays to which the
majority of ongoing assessments have been subject. Instead of
significantly streamlining the process, which GAO recommended, EPA has
institutionalized a process that from the outset is estimated to take 6
to 8 years to complete. This is problematic because of the substantial
rework such cases often require to take into account changing science
and methodologies. Since EPA‘s new process is not responsive to GAO‘s
recommendations, the viability of this critical database has been
further jeopardized.
To view the full product, including the scope and methodology, click on
[hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO-08-810T]. For more
information, contact John B. Stephenson at (202) 512-3841 or
stephensonj@gao.gov.
[End of section]
Mr. Chairman and Members of the Subcommittee:
I am pleased to be here today to discuss issues associated with the
Environmental Protection Agency's (EPA) Integrated Risk Information
System (IRIS)--one of the most significant tools that EPA has developed
to effectively support its mission of protecting people and the
environment from harmful chemical exposures. IRIS contains EPA's
scientific position on the potential human health effects that may
result from exposure to more than 540 chemicals in the environment and
is a critical component of EPA's capacity to support scientifically
sound risk management decisions, policies, and regulations. IRIS is
also relied upon by state and local environmental programs and some
international regulatory bodies for managing their environmental
protection programs. As shown in figure 1, the toxicity assessments in
the IRIS database fulfill the first two critical steps of the four-step
risk assessment process--providing hazard identification and
quantitative dose-response assessments. IRIS information can then be
used with the results of exposure assessments (typically conducted by
EPA's program or regional offices) to provide an overall
characterization of the public health risks for a given chemical in a
given situation. The development of health risk assessments is thus
directly dependent on the development of toxicity assessments such as
those developed in the IRIS program.
Figure 1: National Academies' Risk Assessment and Risk Management Model
Used by EPA:
[See PDF for image]
This figure is an illustration of the National Academies' Risk
Assessment and Risk Management Model Used by EPA. The following
information is depicted:
Risk assessment:
IRIS toxicity assessment:
1. Hazard identification;
2. Dose-response assessments;
3. Exposure assessment;
These three above combine to develop:
4. Risk characterization.
Risk management:
Development of regulatory options;
Evaluation of public health, economic, social, and political
consequences of regulatory options;
The IRIS toxicity assessment combines with the risk management factors
to yield:
Agency decisions and actions.
Source: National Academies.
[End of figure]
Under the National Academies' risk assessment and risk management
paradigm, policy considerations are relevant in the risk management
phase, which occurs after the risk assessment phase.[Footnote 1] With
risk assessment information, decision makers can make informed risk
management decisions on how to protect public health, reflecting other
important data and considerations, such as the costs and benefits of
mitigating identified risks, the technological feasibility of managing
risks, and the concerns of various stakeholders. Examples of risk
management decisions include deciding how much of a chemical a company
may discharge into a river, determining the extent to which a hazardous
waste site must be cleaned up, and setting allowable levels of
contamination in drinking water.
Thus, although IRIS assessments are not regulatory in nature, the
quantitative IRIS values may influence many risk management decisions
and serve as a basis for regulatory consideration. However, EPA's
productivity in finalizing IRIS assessments is poor, and EPA has a
significant backlog of incomplete IRIS assessments and a growing number
of outdated assessments. Importantly, EPA has not been able to complete
assessments of key chemicals of concern to public health, including
dioxin, formaldehyde, trichloroethylene (TCE), naphthalene, and
tetrachloroethylene (perc) (see app. I).
In this context, my testimony today discusses (1) highlights of our
March 2008 report, Chemical Assessments: Low Productivity and New
Interagency Review Process Limit the Usefulness and Credibility of
EPA's Integrated Risk Information System[Footnote 2] and (2) key
aspects of EPA's revised IRIS assessment process, released on April 10,
2008. For our March 2008 report, we examined the outcome of steps EPA
has taken to ensure that IRIS contains current, credible chemical risk
information; to address the backlog of ongoing assessments; and to
respond to new requirements from the Office of Management and Budget
(OMB). We also examined the potential effects of planned changes to the
IRIS assessment process on EPA's ability to ensure that IRIS provides
current, credible risk information. In conducting our work, we obtained
and analyzed information on EPA's productivity and the resources
provided to the program for fiscal years 2000 through 2007, user needs,
and EPA's assessment completion goals. We also interviewed EPA's
National Center for Environmental Assessment officials who manage the
IRIS assessment program; officials from other EPA program offices and
federal science and health agencies involved in the IRIS assessment
process; and officials from the Department of Defense, the Department
of Energy (DOE), the National Aeronautics and Space Administration
(NASA), and OMB. For this testimony, we supplemented our report with a
review of the IRIS assessment process that EPA released on April 10,
2008. We conducted this work from May 7 to May 21, 2008, in accordance
with generally accepted government auditing standards. Those standards
require that we plan and perform the audit to obtain sufficient,
appropriate evidence to provide a reasonable basis for our findings and
conclusions based on our audit objectives. We believe that the evidence
obtained provides a reasonable basis for our findings and conclusions
based on our audit objectives.
Background:
IRIS was created in 1985 to help EPA develop consensus opinions within
the agency about the health effects of chronic exposure to chemicals.
Its importance has increased over time as EPA program offices and the
states have increasingly relied on IRIS information in making
environmental protection decisions. Currently, the IRIS database
contains assessments of more than 540 chemicals. According to EPA,
national and international users access the IRIS database approximately
9 million times a year. EPA's Assistant Administrator for the Office of
Research and Development has described IRIS as the premier national and
international source for qualitative and quantitative chemical risk
information; other federal agencies have noted that IRIS data are
widely accepted by all levels of government across the country for
application of public health policy, providing benefits such as
uniform, standardized methods for toxicology testing and risk
assessment, as well as uniform toxicity values. Similarly, a private-
sector risk assessment expert has stated that the IRIS database has
become the most important source of regulatory toxicity values for use
across EPA's programs and is also widely used across state programs and
internationally.
Historically and currently, the focus of IRIS toxicity assessments has
been on the potential health effects of long-term (chronic) exposure to
chemicals. According to OMB, EPA is the only federal agency that
develops qualitative and quantitative assessments of both cancer and
noncancer risks of exposure to chemicals, and EPA does so largely under
the IRIS program. Other federal agencies develop quantitative estimates
of noncancer effects or qualitative cancer assessments of exposure to
chemicals in the environment. While these latter assessments provide
information on the effects of long-term exposures to chemicals, they
provide only qualitative assessments of cancer risks (known human
carcinogen, likely human carcinogen, etc.) and not quantitative
estimates of cancer potency, which are required to conduct quantitative
risk assessments.
EPA's IRIS assessment process has undergone a number of formal and
informal changes during the past several years. While the process used
to develop IRIS chemical assessments includes numerous individual steps
or activities, major assessment steps include (1) a review of the
scientific literature; (2) preparation of a draft IRIS assessment; (3)
internal EPA reviews of draft assessments; (4) two OMB/interagency
reviews, managed by OMB, that provide input from OMB as well as from
other federal agencies, including those that may be affected by the
IRIS assessments if they lead to regulatory or other actions; (5) an
independent peer review conducted by a panel of experts; and (6) the
completion of a final assessment that is posted to the IRIS Web site.
Unlike many other EPA programs that have statutory requirements,
including specific time frames for completing mandated tasks, the IRIS
program is not subject to statutory requirements or timeframes. In
contrast, the Department of Human Health and Services' Agency for Toxic
Substances and Disease Registry (ATSDR), which develops quantitative
estimates of the noncancer effects of exposures to chemicals in the
environment, is statutorily required to complete its assessments within
certain timeframes.
Findings and Recommendations from Our March 2008 Report on the
Productivity and Credibility of EPA's Integrated Risk Information
System:
The IRIS database is at serious risk of becoming obsolete because the
agency has not been able to routinely complete timely, credible
assessments or decrease a backlog of 70 ongoing assessments.
Specifically, although EPA has taken important steps to improve the
IRIS program and productivity since 2000 and has developed a number of
draft assessments for external review, its efforts to finalize the
assessments have been thwarted by a combination of factors including
the imposition of external requirements, the growing complexity and
scope of risk assessments, and certain EPA management decisions. In
addition, the changes to the IRIS assessment process that EPA was
considering at the time of our review would have added to the already
unacceptable level of delays in completing IRIS assessments and further
limited the credibility of the assessments.
EPA's Efforts to Improve the IRIS Assessment Program Have Not Produced
the Desired Results:
EPA has taken a number of steps to help ensure that IRIS contains
current, credible chemical risk information; to address its backlog of
ongoing assessments; and to respond to new OMB requirements. However,
to date, these changes--including increasing funding, centralizing
staff conducting assessments, and revising the assessment process--have
not enabled EPA to routinely complete credible IRIS assessments or
decrease the backlog. That is, although EPA sent 32 draft assessments
for external review in fiscal years 2006 and 2007, the agency finalized
only 4 IRIS assessments during this time (see fig. 2).
Figure 2: Number of Completed IRIS Assessments, Draft Assessments to
OMB, and IRIS Staff in Full-Time Equivalents, Fiscal Years 2000-2007:
[See PDF for image]
This figure is a multiple line graph depicting the following data:
Fiscal year: 2000;
IRIS staff levels in full-time equivalents: 7.8;
IRIS assessments completed: 5.
Fiscal year: 2001;
IRIS staff levels in full-time equivalents: 7.8;
IRIS assessments completed: 5.
Fiscal year: 2002;
IRIS staff levels in full-time equivalents: 7.8;
IRIS assessments completed: 3.
Fiscal year: 2003;
IRIS staff levels in full-time equivalents: 12.8;
IRIS assessments completed: 11.
Fiscal year: 2004 (Year OMB/Interagency review initiated);
IRIS staff levels in full-time equivalents: 27;
IRIS assessments completed: 4;
Draft IRIS assessments to OMB for OMB/interagency review: 1.
Fiscal year: 2005;
IRIS staff levels in full-time equivalents: 37;
IRIS assessments completed: 4;
Draft IRIS assessments to OMB for OMB/interagency review: 5.
Fiscal year: 2006;
IRIS staff levels in full-time equivalents: 37;
IRIS assessments completed: 2;
Draft IRIS assessments to OMB for OMB/interagency review: 16.
Fiscal year: 2007;
IRIS staff levels in full-time equivalents: 37;
IRIS assessments completed: 2;
Draft IRIS assessments to OMB for OMB/interagency review: 16.
Source: GAO analysis of EPA data.
[End of figure]
Several key factors have contributed to EPA's inability to achieve a
level of productivity that is needed to sustain the IRIS program and
database: new OMB-required reviews of IRIS assessments by OMB and other
federal agencies; the growing complexity and scope of risk assessments;
certain EPA management decisions and issues, including delaying
completion of some assessments to await new research or to develop
enhanced analyses of uncertainty in the assessments; and the
compounding effect of delays. Regarding the last factor, even a single
delay in the assessment process can lead to the need to essentially
repeat the assessment process to take into account changes in science
and methodologies.
A variety of delays have impacted the majority of the 70 assessments
being conducted as of December 2007--48 had been in process for more
than 5 years, and 12 of those for more than 9 years. These time frames
are problematic because of the substantial rework such cases often
require to take into account changing science and methodologies before
they can be completed. For example, EPA's assessment of the cancer
risks stemming from exposure to naphthalene--a chemical used in jet
fuel and in the production of widely used commercial products such as
moth balls, dyes, insecticides, and plasticizers--was nearing
completion in 2006. However, prior to finalizing this assessment, which
had been ongoing for over 4 years, EPA decided that the existing
noncancer assessment had become outdated and essentially restarted the
assessment to include both cancer and noncancer effects. As a result, 6
years after the naphthalene assessment began, it is now back at the
drafting stage. The assessment now will need to reflect relevant
research completed since the draft underwent initial external peer
review in 2004, and it will have to undergo all of the IRIS assessment
steps again, including the additional internal and external reviews
that are now required (see app. I).
Further, because EPA staff time continues to be dedicated to completing
assessments in the backlog, EPA's ability to both keep the more than
540 existing assessments up to date and initiate new assessments is
limited. Importantly, EPA program offices and state and local entities
have requested assessments of hundreds of chemicals not yet in IRIS,
and EPA data as of 2003 indicated that the assessments of 287 chemicals
in the database may be outdated--that is, new information could change
the risk estimates currently in IRIS or enable EPA to develop
additional risk estimates for chemicals in the database (for example,
developing a cancer potency estimate for assessments with only
noncancer estimates). In addition, because EPA's 2003 data are now more
than 4 years old, it is likely that more assessments may be outdated
now.
The consequences of not having current, credible IRIS information can
be significant. EPA's inability to complete its assessment of
formaldehyde, which the agency initiated in 1997 to update information
already in IRIS on the chemical, has had a significant impact on EPA's
air toxics program. Although in 2003 and 2004, the National Cancer
Institute and the National Institute of Occupational Safety and Health
(NIOSH) had released updates to major epidemiological studies of
industrial workers that showed a relationship between formaldehyde and
certain cancers, including leukemia, EPA did not move forward to
finalize an IRIS assessment incorporating these important data.
Instead, EPA opted to await the results of another update to the
National Cancer Institute study. While this additional research was
originally estimated to take, at most, 18 months to complete, at the
time of our report (more than 3 years later) the update was not
complete. In the absence of this information, EPA's Office of Air and
Radiation decided to use risk information developed by an industry-
funded organization--the CIIT Centers for Health Research--for a
national emissions standard. This decision was a factor in EPA
exempting certain facilities with formaldehyde emissions from the
national emissions standard. The CIIT risk estimate indicates a potency
about 2,400 times lower than the estimate in IRIS that was being re-
evaluated and that did not yet consider the 2003 and 2004 National
Cancer Institute and NIOSH epidemiological studies. According to an EPA
official, an IRIS cancer risk factor based on the 2003 and 2004
National Cancer Institute and NIOSH studies would likely be close to
the current IRIS assessment, which EPA has been re-evaluating since
1997. The discrepancy between these two risk estimates raises concerns
about whether the public health is adequately protected in the absence
of current IRIS information. For example, in 1999, EPA published a
national assessment that provided information about the types and
amounts of air toxics to which people are exposed. The assessment,
which also used the CIIT risk estimate for formaldehyde, concluded, for
example, that formaldehyde did not contribute significantly to the
overall cancer risk in the state of New Jersey. However, in carrying
out its own risk assessment on formaldehyde, the New Jersey Department
of Environmental Protection opted to use the risk information that is
currently in IRIS (dating back to 1991) and found that the contribution
from formaldehyde to overall cancer risk in New Jersey is quite
significant, second only to diesel particulate matter. (App. I provides
additional information on EPA's IRIS assessment for formaldehyde.)
One of the factors that has contributed to EPA's inability to complete
assessments in a timely manner--the new OMB-directed OMB/interagency
review process--also limits the credibility of the assessments because
it lacks transparency. Specifically, neither the comments nor the
changes EPA makes to the scientific IRIS assessments in response to the
comments made by OMB and other federal agencies, including those whose
workload and resource levels could be affected by the assessments, are
disclosed. In addition, the OMB/interagency reviews have hindered EPA's
ability to independently manage its IRIS assessments. For example,
without communicating its rationale for doing so, OMB directed EPA to
terminate five IRIS assessments that for the first time addressed
acute, rather than chronic exposure--even though EPA initiated this
type of assessment to help it implement the Clean Air Act.
The Expansion of Agencies' Roles in IRIS Assessments That EPA Was
Considering at the Time of Our Review Would Have Caused Further Delays
and Limited the Assessments' Credibility:
For our March 2008 report, we reviewed the additional assessment
process changes EPA was planning and concluded that they would likely
exacerbate delays in completing IRIS assessments and further affect
their credibility. Specifically, despite the OMB/interagency review
process that OMB required EPA to incorporate into the IRIS assessment
process in 2005, certain federal agencies continued to believe they
should have greater and more formal roles in EPA's development of IRIS
assessments. Consequently, EPA had been working for several years to
establish a formal IRIS assessment process that would further expand
the role of federal agencies in the process--including agencies such as
DOD, which could be affected by the outcome of IRIS assessments. For
example, some of these agencies and their contractors could face
increased cleanup costs and other legal liabilities if EPA issued an
IRIS assessment for a chemical that resulted in a decision to regulate
the chemical to protect the public. In addition, the agencies could be
required to, for example, redesign systems and processes to eliminate
hazardous materials; develop material substitutes; and improve personal
protective clothing, equipment, and procedures. Under the changes that
EPA was planning at the time of our review, these potentially affected
agencies would have the opportunity to be involved, or provide some
form of input, at almost every step of EPA's IRIS assessment process.
Most significantly, the changes would have provided federal agencies,
including those facing potential regulatory liability, with several
opportunities during the IRIS assessment process to subject particular
chemicals of interest to additional process steps. These additional
process steps, which would have lengthened assessment times
considerably, include:
* giving federal agencies and the public 45 days to identify additional
information on a chemical for EPA's consideration in its assessment or
to correct any errors on an additional assessment draft that would
provide qualitative information;[Footnote 3]
* giving potentially affected federal agencies 30 days to review the
public comments EPA received and initiate a meeting with EPA if they
want to discuss a particular set of comments;
* allowing potentially affected federal agencies to have assessments
suspended for up to 18 months to fill a data gap or eliminate an
uncertainty factor that EPA plans to use in its assessment; and:
* allowing other federal agencies to weigh in on (1) the level of
independent peer review that would be sought (that is, whether the peer
reviews would be conducted by EPA Science Advisory Board panels,
National Academies' panels, or panels organized by an EPA contractor);
(2) the areas of scientific expertise needed on the panel; and (3) the
scope of the peer reviews and the specific issues they would address.
EPA estimated that assessments that undergo these additional process
steps would take up to 6 years to complete. While it is important to
ensure that assessments consider the best science, EPA has acknowledged
that waiting for new data can result in substantial harm to human
health, safety, and the environment. Further, although coordination
with other federal agencies about IRIS assessments could enhance their
quality,[Footnote 4] increasing the role of agencies that may be
affected by IRIS assessments in the process itself reduces the
credibility of the assessments if that expanded role is not
transparent. In this regard, while EPA's proposed changes would have
allowed for including federal agencies' comments in the public record,
the implementation of this proposal was delayed for a year, in part,
because of OMB's view that agencies' comments about IRIS assessments
represent internal executive branch communications that may not be made
public--a view that is inconsistent with the principle of sound
science, which relies on, among other things, transparency. (App. II
and III provide flow charts of the IRIS process that was in place at
the time of our review and EPA's draft proposed process being
considered at the time of our review, respectively).
Recommendations Made in Our March 2008 Report:
To address the productivity and credibility issues we identified, we
recommended that the EPA Administrator require the Office of Research
and Development to re-evaluate its draft proposed changes to the IRIS
assessment process in light of the issues raised in our report and
ensure that any revised process, among other things, clearly defines
and documents an IRIS assessment process that will enable the agency to
develop the timely chemical risk information it needs to effectively
conduct its mission. One of our recommendations--that EPA provide at
least 2 years' notice of IRIS assessments that are planned--would,
among other things, provide an efficient alternative to suspending
assessments while waiting for new research because interested parties
would have the opportunity to conduct research before assessments are
started.
In addition, we recommended that the EPA Administrator take steps to
better ensure that EPA has the ability to develop transparent, credible
IRIS assessments--an ability that relies in large part on EPA's
independence in conducting these important assessments. Actions that
are key to this ability include ensuring that EPA can (1) determine the
types of assessments it needs to support EPA programs and (2) define
the appropriate role of external federal agencies in EPA's IRIS
assessment process, and (3) manage an interagency review process in a
manner that enhances the quality, transparency, timeliness, and
credibility of IRIS assessments. In its February 21, 2008, letter
providing comments on our draft report, EPA said it would consider each
of our recommendations in light of the new IRIS process the agency was
developing.
Key Aspects of the Revised IRIS Assessment Process Implemented in April
2008 Which Is Not Responsive to GAO's Recommendations:
On April 10, 2008, EPA issued a revised IRIS assessment process,
effective immediately. Overall, EPA's revised process is not responsive
to the recommendations made in our March 2008 report--it is largely the
same as the draft proposed process we evaluated in our March 2008
report (see app. III and IV). Moreover, changes EPA did incorporate
into the final process are likely to further exacerbate the
productivity and credibility issues we identified in our report.
* We recommended that EPA ensure that, among other things, any revised
process clearly defines and documents a streamlined IRIS assessment
process that can be conducted within time frames that minimize the need
for wasteful rework.
As discussed in our report, when assessments take longer than 2 years,
they can become subject to substantial delays stemming from the need to
redo key analyses to take into account changing science and assessment
methodologies. However, EPA's revised process institutionalizes a
process that the agency estimates will take up to 6 years to complete.
Further, the estimated time frames do not factor in the time for peer
reviews conducted by the National Academies, which can take 2 years to
plan and complete.[Footnote 5] EPA typically uses reviews by the
National Academies for highly controversial chemicals or complex
assessments. Therefore, assessments of key chemicals of concern to
public health that are reviewed by the National Academies are likely to
take at least 8 years to complete. These time frames must also be
considered in light of OMB's view that health assessment values in IRIS
are out of date if they are more than 10 years old and if new
scientific information exists that could change the health assessment
values. Thus, EPA's new process institutionalizes time frames that
could essentially require the agency to start assessment updates as
soon as 2 years after assessments are finalized in order to keep the
IRIS database current. Such time frames are not consistent with our
recommendation that EPA develop, clearly define, and document a
streamlined IRIS process that can be conducted within time frames that
minimize the need for wasteful rework. Further, the agency would need a
significant increase in resources to support such an assessment cycle.
In addition, EPA had previously emphasized that, in suspending
assessments to allow agencies to fill in data gaps, it would allow no
more than 18 months to complete the studies and have them peer
reviewed. However, under the new process, EPA states that it generally
will allow no more than 18 months to complete the studies and have them
peer reviewed. As we concluded in our report, we believe the ability to
suspend assessments for up to 18 months would add to the already
unacceptable level of delays in completing IRIS assessments. Further,
we and several agency officials with whom we spoke believe that the
time needed to plan, conduct, and complete research that would address
significant data gaps, and have it peer reviewed, would likely exceed
18 months. Therefore, the less rigid time frame EPA included in its new
process could result in additional delays.
Finally, the new process expands the scope of one of the additional
steps that initially was to apply only to chemicals of particular
interest to federal agencies.[Footnote 6] Specifically, under the draft
process we reviewed, EPA would have provided an additional review and
comment opportunity for federal agencies and the public for what EPA
officials said would be a small group of chemicals. However, under
EPA's new process, this additional step has been added to the
assessment process for all chemicals and, therefore, will add time to
the already lengthy assessments of all chemicals.
* We also recommended that the EPA Administrator take steps to better
ensure that EPA has the ability to develop transparent, credible IRIS
assessments--an ability that relies in large part on EPA's independence
in conducting these important assessments.
Contrary to our recommendation, EPA has formalized a revised IRIS
process that is selectively, rather than fully, transparent, limiting
the credibility of the assessments. Specifically, while the draft
process we reviewed provided that comments on IRIS assessments from OMB
and other federal agencies would be part of the public record, under
the recently implemented process, comments from federal agencies are
expressly defined as "deliberative" and will not be included in the
public record.[Footnote 7] Given the importance and sensitivity of IRIS
assessments, we believe it is critical that input from all parties,
particularly agencies that may be affected by the outcome of IRIS
assessments, be publicly available. However, under EPA's new process,
input from some IRIS assessment reviewers--representatives of federal
agencies, including those facing potential regulatory liability, and
private stakeholders associated with these agencies--will continue to
receive less public scrutiny than comments from all others.
In commenting on a draft of our March 2008 report, and in a recent
congressional hearing, EPA's Assistant Administrator, Office of
Research and Development, stated that the IRIS process is transparent
because all final IRIS assessments must undergo public and external
peer review. However, as we stated in our report, the presence of
transparency at a later stage of IRIS assessment development does not
explain or excuse its absence earlier. Under the new process, neither
peer reviewers nor the public are privy to the changes EPA makes in
response to the comments OMB and other federal agencies provide to EPA
at several stages in the assessment process--changes to draft
assessments or to the questions EPA poses to the peer review panels.
Importantly, the first IRIS assessment draft that is released to peer
reviewers and to the public includes the undisclosed input from federal
agencies potentially subject to regulation and therefore with an
interest in minimizing the impacts of IRIS assessments on their budgets
and operations.
In addition, EPA's revised process does not provide EPA with sufficient
independence in developing IRIS assessments to ensure they are credible
and transparent. We made several recommendations aimed at restoring
EPA's independence. For example, we recommended that the EPA
Administrator ensure that EPA has the ability to, among other things,
define the appropriate role of external federal agencies in the IRIS
assessment process and determine when interagency issues have been
appropriately addressed. However, under the newly implemented IRIS
assessment process, OMB continues to inform EPA when EPA has adequately
addressed OMB's and interagency comments. This determination must be
made both before EPA can provide draft assessments to external peer
reviewers and to the public and before EPA can finalize and post
assessments on the IRIS database. While EPA officials state that
ultimately IRIS assessments reflect EPA decisions, the new process does
not support this assertion given the clearances EPA needs to receive
from OMB to move forward at key stages. In fact, we believe the new
IRIS assessment process may elevate the goal of reaching interagency
agreement above achieving IRIS program objectives. Further, as
discussed above, because the negotiations over OMB/interagency comments
are not disclosed, whether EPA is entirely responsible for the content
of information on IRIS is open to question.
In our report, we also emphasized the importance of ensuring that IRIS
assessments be based solely on science issues and not policy concerns.
However, under the new IRIS assessment process, EPA has further
introduced policy considerations into the IRIS assessment process. That
is, the newly implemented IRIS assessment process broadens EPA's
characterization of IRIS assessments from "the agency's scientific
positions on human health effects that may result from exposure to
environmental contaminants" to "the agency's science and science policy
positions" on such effects. EPA's new, broader characterization of IRIS
raises concerns about the agency's stated intent to ensure that
scientific assessments are appropriately based on the best available
science and that they are not inappropriately impacted by policy issues
and considerations. For example, in discussing science and science
policy at a recent Senate hearing, EPA's Assistant Administrator of
Research and Development described science policy considerations as
including decisions about filling knowledge gaps (e.g., whether and to
what extent to use default assumptions) and assessing weight-of-the-
evidence approaches to make scientific inferences or assumptions. We
believe that these are scientific decisions that should reflect the
best judgment of EPA scientists who are evaluating the data, using the
detailed risk assessment guidance the agency has developed for such
purposes. We have concerns about the manner and extent to which other
federal agencies, including those that may be affected by the outcome
of assessments, are involved in these decisions as well as the lack of
transparency of their input. As we highlighted earlier, under the
National Academies' risk assessment and risk management paradigm,
policy considerations are relevant in the risk management phase--which
occurs after the risk assessment phase that encompasses IRIS
assessments. The National Academies recently addressed this issue as
follows: "The committee believes that risk assessors and risk managers
should talk with each other; that is, a 'conceptual distinction' does
not mean establishing a wall between risk assessors and risk managers.
Indeed they should have constant interaction. However, the dialogue
should not bias or otherwise color the risk assessment conducted, and
the activities should remain distinct; that is, risk assessors should
not be performing risk management activities."[Footnote 8]
Concluding Observations:
The new IRIS assessment process that EPA implemented in April 2008 will
not allow the agency to routinely and timely complete credible
assessments. In fact, it will exacerbate the problems we identified in
our March 2008 report and sought to address with our recommendations--
all of which were aimed at preserving the viability of this critical
database, which is integral to EPA's mission of protecting the public
and the environment from exposure to toxic chemicals. Specifically,
under the new process, assessment time frames will be significantly
lengthened, and the lack of transparency will further limit the
credibility of the assessments because input from OMB and other
agencies at all stages of the IRIS assessment process is now expressly
defined as deliberative and therefore not subject to public disclosure.
The position of the Assistant Administrator, Office of Research and
Development, that the IRIS process is transparent because all final
IRIS assessments must undergo public and external peer review is
unconvincing. Transparency at a later stage of the IRIS assessment
process--after OMB and other federal agencies have had multiple
opportunities to influence the content of the assessment without any
disclosure of their input--does not compensate for its absence earlier.
We continue to believe that to effectively maintain IRIS EPA must
streamline its lengthy assessment process and adopt transparency
practices that provide assurance that IRIS assessments are
appropriately based on the best available science and that they are not
inappropriately biased by policy issues and considerations. As
discussed in our April 29, 2008, testimony before the Senate
Environment and Public Works Committee, we believe that the Congress
should consider requiring EPA to suspend implementation of its new IRIS
assessment process and develop a streamlined process that is
transparent and otherwise responsive to our recommendations aimed at
improving the timeliness and credibility of IRIS assessments.[Footnote
9] For example, suspending assessments to obtain additional research is
inefficient; alternatively, with longer-term planning, EPA could
provide agencies and the public with more advance notice of
assessments, enabling them to complete relevant research before IRIS
assessments are started.
In addition, as discussed in our April 2008 testimony, the Congress
should consider requiring EPA to obtain and be responsive to input from
the Congress and the public before finalizing a revised IRIS assessment
process. We note that while EPA and OMB initially had planned for EPA
to release a draft revised IRIS assessment process to the public, hold
a public meeting to discuss EPA's proposed changes, and seek and
incorporate public input before finalizing the process, EPA released
its new assessment process without obtaining public input and made it
effective immediately. This was inconsistent with assertions made in
OMB's letter commenting on our draft report, which emphasized that EPA
had not completed the development of the IRIS assessment process and
stated: "Indeed, the process will not be complete until EPA circulates
its draft to the public for comments and then releases a final product
that is responsive to those comments.":
Finally, if EPA is not able to take the steps we have recommended to
effectively maintain this critical program, other approaches, including
statutory requirements, may need to be explored.
Mr. Chairman, this concludes my prepared statement. I would be happy to
respond to any questions that you or other Members of the Subcommittee
may have at this time.
Contacts and Acknowledgments:
For further information about this testimony, please contact John B.
Stephenson on (202) 512-3841 or stephensonj@gao.gov. Contact points for
our Congressional Relations and Public Affairs Offices may be found on
the last page of this statement. Contributors to this testimony include
Christine Fishkin (Assistant Director), Laura Gatz, Richard P. Johnson,
and Nancy Crothers.
[End of section]
Appendix I: Examples of Key IRIS Assessments That Have Been Delayed:
Some key IRIS assessments have been in progress for a number of years,
in part because of delays stemming from one or more of the key factors
we identified that have hindered EPA's productivity.[Footnote 10]
Examples include the following:
Naphthalene. EPA started the IRIS assessment of cancer risks stemming
from the inhalation of naphthalene in 2002. Naphthalene is used in jet
fuel and in the production of widely used commercial products such as
moth balls, dyes, insecticides, and plasticizers. According to a
presentation delivered at the 2007 annual meeting of the Society for
Risk Analysis by an Army Corps of Engineers toxicologist,[Footnote 11]
"The changing naphthalene regulatory environment includes a draft EPA
risk assessment that if/when finalized, will change naphthalene's
status from 'possible' to 'likely' human carcinogen."[Footnote 12]
Thus, according to this presentation, one potential impact of this IRIS
assessment on DOD is that DOD would need to provide many employees
exposed to naphthalene with equipment measuring their exposure to the
chemical. In addition, because many military bases are contaminated
with naphthalene, a component of jet fuel (approximately 1 percent to 3
percent) used by all DOD services, DOD could face extensive cleanup
costs. By 2004, 2 years after starting the assessment, EPA had drafted
a chemical assessment that had completed internal peer reviews and was
about to be sent to an external peer review committee. Once it returned
from external review, the next step, at that time, would have been a
formal review by EPA's IRIS Agency Review Committee. If approved, the
assessment would have been completed and released. However, in part
because of concerns raised by DOD, OMB asked to review the assessment
and conducted an interagency review of the draft. In their 2004 reviews
of the draft IRIS assessment, both OMB and DOD raised a number of
concerns about the assessment and suggested to EPA that it be suspended
until additional research could be completed to address what they
considered to be significant uncertainties associated with the
assessment. Although all of the issues raised by OMB and DOD were not
resolved, EPA continued with its assessment by submitting the draft for
external peer review, which was completed in September 2004.[Footnote
13] However, according to EPA, OMB continued to object to the draft
IRIS assessment and directed EPA to convene an additional expert review
panel on genotoxicity to obtain recommendations about short-term tests
that OMB thought could be done quickly.[Footnote 14] According to EPA,
this added 6 months to the process, and the panel, which met in April
2005, concluded that the research that OMB was proposing could not be
conducted in the short term. Nonetheless, EPA officials said that the
second expert panel review did not eliminate OMB's concerns regarding
the assessment, which they described as reaching a stalemate. In
September 2006, EPA decided, however, to proceed with developing the
assessment. By this time, the naphthalene assessment had been in
progress for over 4 years; EPA decided that the IRIS noncancer
assessment, issued in 1998, was outdated and needed to be revisited.
Thus, EPA expanded the IRIS naphthalene assessment to include both
noncancer and cancer assessments. As a result, 6 years after the
naphthalene assessment began, it is now back at the drafting stage. The
assessment now will need to reflect relevant research completed since
the draft underwent initial external peer review in 2004, and it will
have to undergo all of the IRIS assessment steps again, including
additional internal and external reviews that are now required. This
series of delays has limited EPA's ability to conduct its mission. For
example, the Office of Air and Radiation has identified the naphthalene
assessment as one of its highest-priority needs for its air toxics
program. In addition, the Office of Solid Waste and Emergency Response
considers the naphthalene assessment a high priority for the Superfund
program--naphthalene has been found in at least 654 of Superfund's
current or former National Priorities List sites.[Footnote 15] Although
EPA currently estimates that it will complete the assessment in 2009,
meeting this revised estimate will be challenging, given all of the
steps that are yet to be completed and the extensive external scrutiny
to which it will continue to be subjected.
Royal Demolition Explosive. This chemical, also called RDX or hexahydro-
1,3,5-trinitrotriazine, is a highly powerful explosive used by the U.S.
military in thousands of munitions. Currently classified by EPA as a
possible human carcinogen, this chemical is known to leach from soil to
groundwater. Royal Demolition Explosive can cause seizures in humans
and animals when large amounts are inhaled or ingested, but the effects
of long-term, low-level exposure on the nervous system are unknown. As
is the case with naphthalene, the IRIS assessment could potentially
require DOD to undertake a number of actions, including steps to
protect its employees from the effects of this chemical and to clean up
many contaminated sites. Although EPA started an IRIS assessment of
Royal Demolition Explosive in 2000, it has made minimal progress on the
assessment because EPA agreed to a request by DOD to wait for the
results of DOD-sponsored research on this chemical. In 2007, EPA began
to actively work on this assessment, although some of the DOD-sponsored
research is still outstanding.
Formaldehyde. EPA began an IRIS assessment of formaldehyde in 1997
because the existing assessment was determined to be outdated.[Footnote
16] Formaldehyde is a colorless, flammable, strong-smelling gas used to
manufacture building materials, such as pressed wood products, and used
in many household products, including paper, pharmaceuticals, and
leather goods. While EPA currently classifies formaldehyde as a
probable human carcinogen, the International Agency for Research on
Cancer (IARC), part of the World Health Organization, classifies
formaldehyde as a known human carcinogen. Since 1986, studies of
industrial of workers have suggested that formaldehyde exposure is
associated with nasopharyngeal cancer, and possibly with leukemia. For
example, in 2003 and 2004, the National Cancer Institute (NCI) and the
National Institute of Occupational Safety and Health (NIOSH) released
epidemiological studies following up on earlier studies tracking about
26,000 and 11,000 industrial workers, respectively, exposed to
formaldehyde; the updates showed exposure to formaldehyde might also
cause leukemia in humans, in addition to the cancer types previously
identified. According to NCI officials, the key findings in their
follow-up study were an increase in leukemia deaths and, more
significantly, an exposure/response relationship between formaldehyde
and leukemia--as exposure increased, the incidence of leukemia also
rose. As with the earlier study, NCI found more cases of a rare form of
cancer, nasopharyngeal cancer, than would usually be expected. The
studies from NCI and NIOSH were published in 2003 and 2004,[Footnote
17] around the time that EPA was still drafting its IRIS assessment. In
November 2004, the Chairman of the Senate Environment and Public Works
Committee requested that EPA delay completion of its IRIS assessment
until an update to the just-released NCI study could be conducted,
indicating that the effort would take, at most, 18 months. EPA agreed
to wait--and more than 3 years later, the NCI update is not yet
complete. As of December 2007, NCI estimates that the study will be
completed in two stages, one in mid-2008 and the second one later that
year. An NCI official said that the additional leukemia deaths
identified in the update provide "greater power" to detect associations
between exposure to formaldehyde and cancer. EPA's inability to
complete the IRIS assessment it started more than 10 years ago in a
timely manner has had a significant impact on EPA's air toxics program.
Specifically, when EPA promulgated a national emissions standard for
hazardous air pollutants covering facilities in the plywood and
composite wood industries in 2004, EPA's Office of Air and Radiation
took the unusual step of not using the existing IRIS estimate but
rather decided to use a cancer risk estimate developed by an industry-
funded organization, the CIIT Centers for Health Research (formerly,
the Chemical Industry Institute of Toxicology) that had been used by
the Canadian health protection agency. The IRIS cancer risk factor had
been subject to criticism because it was last revised in 1991 and was
based on data from the 1980s. In its final rule, EPA stated that "the
dose-response value in IRIS is based on a 1987 study, and no longer
represents the best available science in the peer-reviewed literature."
The CIIT quantitative cancer risk estimate that EPA used in its health
risk assessment in the plywood and composite wood national emissions
standard indicates a potency about 2,400 times lower than the estimate
in IRIS that was being re-evaluated and that did not yet consider the
2003 and 2004 NCI and NIOSH epidemiological studies. According to an
EPA official, an IRIS cancer risk factor based on the 2003 and 2004 NCI
and NIOSH studies would likely be close to the current IRIS assessment,
which EPA has been attempting to update since 1997. The decision to use
the CIIT assessment in the plywood national emissions standard was
controversial, and officials in EPA's National Center for Environmental
Assessment said the center identified numerous problems with the CIIT
estimate. Nonetheless, the Office of Air and Radiation used the CIIT
value, and that decision was a factor in EPA exempting certain
facilities with formaldehyde emissions from the national emissions
standard. In June 2007, a federal appellate court struck down the rule,
holding that EPA's decision to exempt certain facilities that EPA
asserted presented a low health risk exceeded the agency's authority
under the Clean Air Act.[Footnote 18] Further, the continued delays of
the IRIS assessment of formaldehyde--currently estimated to be
completed in 2010 but after almost 11 years still in the draft
development stage--will impact the quality of other EPA regulatory
actions, including other air toxics rules and requirements.
Trichloroethylene. Also known as TCE, this chemical is a solvent widely
used as a degreasing agent in industrial and manufacturing settings; it
is a common environmental contaminant in air, soil, surface water, and
groundwater. TCE has been linked to cancer, including childhood cancer,
and other significant health hazards, such as birth defects. TCE is the
most frequently reported organic contaminant in groundwater, and
contaminated drinking water has been found at Camp Lejeune, a large
Marine Corps base in North Carolina. TCE has also been found at
Superfund sites and at many industrial and government facilities,
including aircraft and spacecraft manufacturing operations. In 1995,
the International Agency for Research on Cancer classified TCE as a
probable human carcinogen, and in 2000, the Department of Health and
Human Services' National Toxicology Program concluded that it is
reasonably anticipated to be a human carcinogen. Because of questions
raised by peer reviewers about the IRIS cancer assessment for TCE, EPA
withdrew it from IRIS in 1989 but did not initiate a new TCE cancer
assessment until 1998. In 2001, EPA issued a draft IRIS assessment for
TCE that proposed a range of toxicity values indicating a higher
potency than in the prior IRIS values and characterizing TCE as "highly
likely to produce cancer in humans." The draft assessment, which became
controversial, was peer reviewed by EPA's Scientific Advisory Board and
released for public comment. A number of scientific issues were raised
during the course of these reviews, including how EPA had applied
emerging risk assessment methods--such as assessing cumulative effects
(of TCE and its metabolites) and using a physiologically based
pharmacokinetic model--and the uncertainty associated with the new
methods themselves.[Footnote 19] To help address these issues, EPA,
DOD, DOE, and NASA sponsored a National Academies review to provide
guidance. The National Academies report, which was issued in 2006,
concluded that the weight of evidence of cancer and other health risks
from TCE exposure had strengthened since 2001 and recommended that the
risk assessment be finalized with currently available data so that risk
management decisions could be made expeditiously. The report
specifically noted that while some additional information would allow
for more precise estimates of risk, this information was not necessary
for developing a credible risk assessment. Nonetheless, 10 years after
EPA started its IRIS assessment, the TCE assessment is back at the
draft development stage. EPA estimates this assessment will be
finalized in 2010. More in line with the National Academies'
recommendation to act expeditiously, five senators introduced a bill in
August 2007 that, among other things, would require EPA to both
establish IRIS values for TCE and issue final drinking water standards
for this contaminant within 18 months.
Tetrachloroethylene. EPA started an IRIS assessment of
tetrachloroethylene--also called perchloroethylene or "perc"--in 1998.
Tetrachloroethylene is a manufactured chemical widely used for dry
cleaning of fabrics, metal degreasing, and making some consumer
products and other chemicals. Tetrachloroethylene is a widespread
groundwater contaminant, and the Department of Health and Human
Services' National Toxicology Program has determined that it is
reasonably anticipated to be a carcinogen. The IRIS database currently
contains a 1988 noncancer assessment based on oral exposure that will
be updated in the ongoing assessment. Importantly, the ongoing
assessment will also provide a noncancer inhalation risk and a cancer
assessment. The IRIS agency review of the draft assessment was
completed in February 2005, the draft assessment was sent to OMB for
OMB/interagency review in September 2005, and the OMB/interagency
review was completed in March 2006. EPA had determined to have the next
step, external peer review, conducted by the National Academies--the
peer review choice reserved for chemical assessments that are
particularly significant or controversial. EPA contracted with the
National Academies for a review by an expert panel, and the review was
scheduled to start in June 2006 and be completed in 15 months. However,
as of December 2007, the draft assessment had not yet been provided to
the National Academies. After verbally agreeing with both the noncancer
and cancer assessments following briefings on the assessments, the
Assistant Administrator, Office of Research and Development,
subsequently requested that additional uncertainty analyses--including
some quantitative analyses--be conducted and included in the assessment
before the draft was released to the National Academies for peer
review. As discussed in our March 2008 report on IRIS [hyperlink,
http://www.gao.gov/cgi-bin/getrpt?GAO-08-440], quantitative uncertainty
analysis is a risk assessment tool that is currently being developed,
and although the agency is working on developing policies and
procedures for uncertainty analysis, such guidance currently does not
exist. The draft tetrachloroethylene assessment has been delayed since
early 2006 as EPA staff have gone back and forth with the Assistant
Administrator trying to reach agreement on key issues such as whether a
linear or nonlinear model is most appropriate for the cancer assessment
and how uncertainty should be qualitatively and quantitatively
characterized. EPA officials and staff noted that some of the most
experienced staff are being used for these efforts, limiting their
ability to work on other IRIS assessments. In addition, the significant
delay has impacted the planned National Academies peer review because
the current contract, which has already been extended once, cannot be
extended beyond December 2008. The peer review was initially estimated
to take 15 months. As a result, a new contract and the appointment of
another panel may be required.
Dioxin. The dioxin assessment is an example of an IRIS assessment that
has been, and will likely continue to be, a political as well as a
scientific issue. Often the byproducts of combustion and other
industrial processes, complex mixtures of dioxins enter the food chain
and human diet through emissions into the air that settle on soil,
plants, and water. EPA's initial dioxin assessment, published in 1985,
focused on the dioxin TCDD (2,3,7,8-tetrachlorodibenzo-p-dioxin)
because animal studies in the 1970s showed it to be the most potent
cancer-causing chemical studied to date. Several years later, EPA
decided to conduct a reassessment of dioxin because of major advances
that had occurred in the scientific understanding of dioxin toxicity
and significant new studies on dioxins' potential adverse health
effects. Initially started in 1991, this assessment has involved
repeated literature searches and peer reviews. For example, a draft of
the updated assessment was reviewed by a scientific peer review panel
in 1995, and three panels reviewed key segments of later versions of
the draft in 1997 and 2000. In 2002, EPA officials said that the
assessment would conclude that dioxin may adversely affect human health
at lower exposure levels than had previously been thought and that most
exposure to dioxins occurs from eating such American dietary staples as
meats, fish, and dairy products, which contain minute traces of
dioxins. These foods contain dioxins because animals eat plants and
commercial feed and drink water contaminated with dioxins, which then
accumulate in animals' fatty tissue. It is clear that EPA's dioxin risk
assessment could have a potentially significant impact on consumers and
on the food and agriculture industries. As EPA moved closer to
finalizing the assessment, in 2003 the agency was directed in a
congressional appropriations conference committee report to not issue
the assessment until it had been reviewed by the National Academies.
The National Academies provided EPA with a report in July 2006. In
developing a response to the report, which the agency is currently
doing, EPA must include new studies and risk assessment approaches that
did not exist when the assessment was drafted. EPA officials said the
assessment will be subject to the IRIS review process once its response
to the National Academies' report is drafted. As of 2008, EPA has been
developing the dioxin assessment, which has potentially significant
health implications for all Americans, for 17 years.
[End of section]
Appendix II: EPA's IRIS Assessment Process Being Implemented at the
Time of Our Review (Includes OMB Requirements as of 2005):
This appendix contains a figure that is an illustration of a flowchart
of EPA's IRIS Assessment Process, as follows:
1) Annual request for nomination of IRIS chemicals for assessment.
2) Determine the annual agenda and publish it in the Federal Register,
along with a call for scientific information from the public on select
chemicals.
3) Scientific literature search.
4) Development of a complete draft IRIS assessment (qualitative and
quantitative), including internal peer consultation.
5) Internal agency review by the 18-member IRIS Agency Review
Committee.
6) Revised draft IRIS assessment.
7) OMB/interagency review coordinated by OMB.
8) Revised draft IRIS assessment.
9) Independent external peer review and public comment.
10) Revised draft IRIS assessment.
11a) Second internal agency review by the 18-member IRIS Agency Review
Committee.
11b) Second OMB/interagency review coordinated by OMB.
[Both 11a and 11b may lead to an additional revised draft IRIS
assessment before the final step.]
12) Completion of IRIS assessment, EPA management review and approval,
and posting on IRIS.
Source: GAO analysis of EPA information.
[End of section]
Appendix III: EPA's Draft Proposed IRIS Assessment Process Being
Considered at the Time of Our Review (Dated March 2007):
This appendix contains a figure that is an illustration of a flowchart
of EPA's Draft Proposed IRIS Assessment Process, as follows:
1) Annual request for nomination of IRIS chemicals for assessment
(step where EPA has provided additional opportunity for input from
potentially affected federal agencies for all chemicals);
2) Determine the annual agenda and publish it in the Federal Register
(step where EPA has provided additional opportunity for input from
potentially affected federal agencies for all chemicals);
3) Scientific literature search;
4) Federal Register notice/data call-in (step where EPA has provided
additional opportunity for input from potentially affected federal
agencies for all chemicals);
5) Development of a draft qualitative assessment;
6) Is the chemical mission critical? If yes, proceed to step 7; if no,
proceed to step 12;
7) Public review of draft qualitative assessment (additional step,
under EPA‘s planned changes, to its assessment process and steps where
EPA has provided additional opportunity for input from potentially
affected federal agencies for mission-critical chemicals);
8) Review of public comments (additional step, under EPA‘s planned
changes, to its assessment process and steps where EPA has provided
additional opportunity for input from potentially affected federal
agencies for mission-critical chemicals);
9) Interagency evaluation to close data gaps (additional step, under
EPA‘s planned changes, to its assessment process and steps where EPA
has provided additional opportunity for input from potentially affected
federal agencies for mission-critical chemicals);
10) Is there interest in conducting research to close data gaps? If
yes, proceed to step 11; if no proceed to step 12;
11) Federal agency development of new studies; (additional step, under
EPA‘s planned changes, to its assessment process and steps where EPA
has provided additional opportunity for input from potentially affected
federal agencies for mission-critical chemicals);
12) Development of a complete draft IRIS assessment (qualitative and
quantitative), including internal peer consultation;
13) Is the chemical mission critical? If yes, proceed to step 14a; if
no, proceed to step 14b;
14a) EPA consults with federal agencies to determine level and scope of
peer review (additional step, under EPA‘s planned changes, to its
assessment process and steps where EPA has provided additional
opportunity for input from potentially affected federal agencies for
mission-critical chemicals);
14b) EPA determines the level and scope of the peer review;
15) Internal agency review by the 18-member IRIS Agency Review
Committee;
16) Revised draft IRIS assessment (may require a return to step 15);
17) OMB/interagency review coordinated by OMB (step where potentially
affected federal agencies already had an opportunity for input);
18) Revised draft IRIS assessment (may require a return to step 17);
19) Independent external peer review and public comment (step where
potentially affected federal agencies already had an opportunity for
input);
20) Revised draft IRIS assessment;
21a) Second internal agency review by the 18-member IRIS Agency Review
Committee (may require a return to step 20);
21b) Second OMB/interagency review coordinated by OMB (may require a
return to step 20); (step where potentially affected federal agencies
already had an opportunity for input);
22) Completion of IRIS assessment, EPA management review and approval,
and posting on IRIS.
Source: GAO analysis of EPA information.
[End of section]
Appendix IV: EPA's IRIS Assessment Process as of April 10, 2008:
This appendix contains a figure that is an illustration of a flowchart
of EPA's IRIS Assessment Process, as follows:
1) Annual request for nomination of IRIS chemicals for assessment
(step where EPA has provided additional opportunity for input from
potentially affected federal agencies for all chemicals);
2) Determine the annual agenda and publish it in the Federal Register
(step where EPA has provided additional opportunity for input from
potentially affected federal agencies for all chemicals);
3) Scientific literature search;
4) Federal Register notice/data call-in (step where EPA has provided
additional opportunity for input from potentially affected federal
agencies for all chemicals);
5) Development of a draft qualitative assessment;
6) Federal agency and public review of draft qualitative assessment
(step where EPA has provided additional opportunity for input from
potentially affected federal agencies for all chemicals);
7) EPA and federal agency review of comments (step where EPA has
provided additional opportunity for input from potentially affected
federal agencies for all chemicals);
8) Is the chemical mission critical? If yes, proceed to step 9; if no
proceed to step 12;
9) Interagency evaluation to close data gaps (additional step, under
EPA‘s planned changes, to its assessment process and steps where EPA
has provided additional opportunity for input from potentially affected
federal agencies for mission-critical chemicals);
10) Is there interest in conducting research to close data gaps? If
yes, proceed to step 11; if no proceed to step 12;
11) Federal agency development of new studies; (additional step, under
EPA‘s planned changes, to its assessment process and steps where EPA
has provided additional opportunity for input from potentially affected
federal agencies for mission-critical chemicals);
12) Development of a complete draft IRIS assessment (qualitative and
quantitative), including internal peer consultation;
13) Is the chemical mission critical? If yes, proceed to step 14a; if
no, proceed to step 14b;
14a) EPA consults with federal agencies to determine level and scope of
peer review (additional step, under EPA‘s planned changes, to its
assessment process and steps where EPA has provided additional
opportunity for input from potentially affected federal agencies for
mission-critical chemicals);
14b) EPA determines the level and scope of the peer review;
15) Internal agency review by the 18-member IRIS Agency Review
Committee;
16) Revised draft IRIS assessment (may require a return to step 15);
17) OMB/interagency review coordinated by OMB (step where potentially
affected federal agencies already had an opportunity for input);
18) Revised draft IRIS assessment (may require a return to step 17);
19) Independent external peer review and public comment (step where
potentially affected federal agencies already had an opportunity for
input);
20) Revised draft IRIS assessment;
21a) Second internal agency review by the 18-member IRIS Agency Review
Committee (may require a return to step 20);
21b) Second OMB/interagency review coordinated by OMB (may require a
return to step 20); (step where potentially affected federal agencies
already had an opportunity for input);
22) Completion of IRIS assessment, EPA management review and approval,
and posting on IRIS.
Source: GAO analysis of EPA information.
[End of section]
Footnotes:
[1] The National Academies comprises four organizations: the National
Academy of Sciences, the National Academy of Engineering, the Institute
of Medicine, and the National Research Council.
[2] [hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO-08-440]
(Washington, D.C.: Mar. 7, 2008).
[3] This represents an additional review of a new draft product and
comment period that had not existed previously. As shown in app. II,
the assessment process EPA used at the time of our review included
publishing its annual IRIS assessment agenda in the Federal Register
and soliciting relevant scientific information from the public.
[4] We recommended in our 2006 report on human health risk assessment
that EPA consistently involve stakeholders as appropriate to the risk
assessment. We made this recommendation in the context of improving the
overall quality, consistency, and transparency of risk assessments.
GAO, Human Health Risk Assessment: EPA Has Taken Steps to Strengthen
Its Process, but Improvements Needed in Planning, Data Development, and
Training, [hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO-06-595]
(Washington, D.C.: May 31, 2006).
[5] It is not clear whether the time frames exclude reviews conducted
by EPA's Science Advisory Board, which can also add considerably more
time than the most basic level of peer review used by the IRIS program-
-panels organized by an EPA contractor.
[6] The new IRIS assessment process refers to such chemicals as
"mission critical." The process defines a mission-critical chemical as
one that "is an integral component to the successful and safe conduct
of an agency's mission in any or all phases of its operations."
According to the process, "impacts on the use of mission-critical
chemicals include cessation or degradation of the conduct of the
mission and/or unacceptable resource constraints."
[7] Making these comments public would have been a change from the OMB/
interagency review process that has been in place since 2004.
[8] National Academies, Scientific Review of the Proposed Risk
Assessment Bulletin from the Office of Management and Budget Committee
to Review the OMB Risk Assessment Bulletin (2007).
[9] GAO, Toxic Chemicals: EPA's New Assessment Process Will Increase
Challenges EPA Faces in Evaluating and Regulating Chemicals,
[hyperlink, http://www.gao.gov/cgi-bin/getrpt?GAO-08-743T] (Washington,
D.C.: April 29, 2008).
[10] The factors we identified that have hindered EPA's efforts to
improve productivity are the OMB/interagency review process managed by
OMB, the growing complexity and scope of risk assessments, certain
management decisions and issues regarding the IRIS program,
congressional action that has delayed some assessments with potentially
significant economic effects, and the compounding effect of delays.
[11] Presentations at the Society for Risk Analysis meeting reflect the
views of the authors and "do not necessarily reflect the views of any
other organization or agency."
[12] Using its 1996 Proposed Guidelines for Carcinogen Risk Assessment,
EPA concluded in the 1998 IRIS assessment of naphthalene that its human
carcinogenic potential could not be determined at that time, but noted
that there was suggestive evidence of potential human carcinogenicity.
(EPA also noted that under its 1986 cancer guidelines, EPA classified
naphthalene as a possible human carcinogen.) Subsequently, in 2002, the
International Agency for Research on Cancer (IARC), part of the World
Health Organization, concluded that naphthalene is possibly
carcinogenic to humans; in 2004, the Department of Human Health and
Services' National Toxicology Program concluded that naphthalene can
reasonably be anticipated to be a human carcinogen. EPA's current
assessment will be subject to the agency's 2005 cancer guidelines.
[13] According to DOD, EPA did not specifically ask the peer reviewers
to address some of the technical questions DOD had raised and wanted
the peer review to address.
[14] Genotoxic substances are a type of carcinogen, specifically those
capable of causing genetic mutation and of contributing to the
development of tumors. This includes both certain chemical compounds
and certain types of radiation.
[15] The National Priorities List is EPA's list of seriously
contaminated sites.
[16] The cancer portion of the formaldehyde assessment was originally
issued in 1989 and updated in 1991; the noncancer assessment was added
in 1990.
[17] NCI published the results of its study in two publications. The
first study, published in November 2003, focused on the association
between formaldehyde exposure and leukemia. M. Hauptmann, J. H. Lubin,
P. A. Stewart, R. B. Hayes, A. Blair, "Mortality from
Lymphohematopoietic Malignancies among Workers in Formaldehyde
Industries," Journal of the National Cancer Institute (2003). The
second study, published in June 2004, evaluated the association between
formaldehyde exposure and other cancers--including nasopharyngeal
cancer. M. Hauptmann, J. H. Lubin, P. A. Stewart, R. B. Hayes, A.
Blair, "Mortality from Solid Cancers among Workers in Formaldehyde
Industries," American Journal of Epidemiology (2004). The results of
the NIOSH study were described in one publication, dated March 2004,
which assessed mortality from all causes and all cancers. L. E.
Pinkerton, M. J. Hein, L. T. Stayner, "Mortality among a Cohort of
Garment Workers Exposed to Formaldehyde: an Update," Occupational and
Environmental Medicine (2004).
[18] Natural Resources Defense Council v. E.P.A., 489 F.3d 1364, 1372-
73 (D.C. Cir, 2007). The court did not specifically address EPA's
reliance on the CIIT study, holding instead that the Clean Air Act
prohibited establishment of the exemptions at issue.
[19] Physiologically based pharmacokinetic models are a class of
dosimetry models that are useful for predicting internal doses to
target organs. With the appropriate data, these models can be used to
extrapolate across species and exposure scenarios and address various
sources of uncertainty in risk assessments.
[End of section]
GAO's Mission:
The Government Accountability Office, the audit, evaluation and
investigative arm of Congress, exists to support Congress in meeting
its constitutional responsibilities and to help improve the performance
and accountability of the federal government for the American people.
GAO examines the use of public funds; evaluates federal programs and
policies; and provides analyses, recommendations, and other assistance
to help Congress make informed oversight, policy, and funding
decisions. GAO's commitment to good government is reflected in its core
values of accountability, integrity, and reliability.
Obtaining Copies of GAO Reports and Testimony:
The fastest and easiest way to obtain copies of GAO documents at no
cost is through GAO's Web site [hyperlink, http://www.gao.gov]. Each
weekday, GAO posts newly released reports, testimony, and
correspondence on its Web site. To have GAO e-mail you a list of newly
posted products every afternoon, go to [hyperlink, http://www.gao.gov]
and select "E-mail Updates."
Order by Mail or Phone:
The first copy of each printed report is free. Additional copies are $2
each. A check or money order should be made out to the Superintendent
of Documents. GAO also accepts VISA and Mastercard. Orders for 100 or
more copies mailed to a single address are discounted 25 percent.
Orders should be sent to:
U.S. Government Accountability Office:
441 G Street NW, Room LM:
Washington, D.C. 20548:
To order by Phone:
Voice: (202) 512-6000:
TDD: (202) 512-2537:
Fax: (202) 512-6061:
To Report Fraud, Waste, and Abuse in Federal Programs:
Contact:
Web site: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]:
E-mail: fraudnet@gao.gov:
Automated answering system: (800) 424-5454 or (202) 512-7470:
Congressional Relations:
Ralph Dawn, Managing Director, dawnr@gao.gov:
(202) 512-4400:
U.S. Government Accountability Office:
441 G Street NW, Room 7125:
Washington, D.C. 20548:
Public Affairs:
Chuck Young, Managing Director, youngc1@gao.gov:
(202) 512-4800:
U.S. Government Accountability Office:
441 G Street NW, Room 7149:
Washington, D.C. 20548:
