REGULATION AND CARCINOGENS: THE PROBLEM AND A SOLUTION
This article also appears in the Oak Ridge National Laboratory Review
(Vol. 26, No. 1), a quarterly research and development magazine. If
you'd like more information about the research discussed in the article
or about the Review, or if you have any helpful comments, drop us a line. Thanks for reading the Review.
A graduate student doing her dissertation on waste disposal examined two
different burial sites for low-level radioactive wastes. The first was
a new site currently receiving waste for disposal, and the second was an
old site currently undergoing cleanup. She was told that the levels of
radioactivity in materials that could be disposed of at the new site are
much greater than the goal for levels of radioactivity after cleanup at
the old site. When she asked about this difference, she was told that
the two sites are regulated under different laws, and the maximum
allowable health risk to the public from disposal at the new site is at
least 100 times greater than the risk goal for cleanup at the old site.
This hypothetical situation involving two radioactive waste sites
illustrates a general problem with laws and regulations for hazardous
substances in the environment. This problem and a possible solution are
discussed in this article.
Routine exposures to radionuclides and other carcinogenic
(cancer-causing) substances in the environment are controlled under a
variety of laws and regulations that essentially limit health risks to
the public. However, clear inconsistencies exist in the levels of health
risk regarded as acceptable for two general categories of standards:
(1) standards that apply only to radionuclides, as developed under the
authority of the Atomic Energy Act, and (2) standards that apply to any
carcinogens, including radionuclides, or only to chemical carcinogens,
as developed under the authority of other laws.
Standards in the first category apply to any radionuclides associated
with the nuclear fuel cycle, such as radionuclides produced in nuclear
power plants. Standards in the second category apply to a wide variety
of other man-made or naturally occurring sources of exposure to
radionuclides and chemical carcinogens. All standards for environmental
carcinogens fall into one of these two categories. Note that the two
categories of standards do not apply exclusively to either radionuclides
or chemical carcinogens because radiation exposures can be regulated in
either category, depending on their source.
In this article, we first discuss the inconsistency in levels of
acceptable health risk for the two categories of standards described
previously. This inconsistency arises from the fundamentally different
approaches to regulation taken in the two cases.
We then propose a set of principles we believe would provide more
consistent regulation of health risks to the public from exposure to any
environmental carcinogens. Such a consistent approach would encourage
consideration of risks from exposure to any chemical carcinogen or
radiation source in the context of the total cancer risk from all
sources and would ensure that cancer risks much less than the largely
unavoidable background risks do not receive unwarranted attention.
Ultimately, a more consistent regulatory approach should save money and
benefit public health by ensuring that the greatest emphasis is placed
on those exposure situations involving the most important risks.
Our proposal for more consistent regulation is based primarily on
distinguishing unambiguously between unacceptable cancer risks from any
source of exposure and risks that are trivial. Our proposal also takes
into account such important factors as the costs and benefits of
reducing risks, technical feasibility, and public perceptions of risk.
TOP-DOWN APPROACH
The framework for regulating routine radiation exposures of the public
under the Atomic Energy Act is what we call a "top-down" approach. This
approach has two components. First, a limit on radiation exposure
corresponding to an upper bound for acceptable risk is established.
Then, exposures are reduced below the limit by requiring all exposures
to be "as low as reasonably achievable" (ALARA). The ALARA principle
takes into account costs and benefits, technical feasibility, and
societal concerns about cancer risks.
The top-down approach is used in radiation protection standards, which
limit the public's total exposure to all sources of man-made
radionuclides associated with the nuclear fuel cycle, which includes
uranium processing and enrichment facilities, nuclear power plants, fuel
reprocessing facilities, and radioactive waste disposal sites. The upper
bound on acceptable risk implicit in these standards is estimated as
follows. The current limit on radiation dose equivalent for members of
the public is 1 millisievert (mSv), or 100 mrem, per year. The
International Commission on Radiological Protection (ICRP) recommends
calculating the increased risk of developing a fatal cancer as a result
of exposure to radiation using a risk factor of 0.05 per Sv. Thus, for
continuous exposure over an average lifetime of 70 years, the dose limit
corresponds to an upper bound on acceptable lifetime risk of about 4 in
1000. In other words, if a population of 1000 individuals were exposed
throughout their lifetimes to the maximum allowable radiation dose, 4 of
them would be expected to die from cancer as a result of this exposure.
However, it is very unlikely that the lifetime risk to members of the
public from routine exposure to all man-made radionuclides could
approach 4 in 1000. The development of many standards that limit doses
from particular practices or sources to levels well below 1 mSv per year
virtually ensures that the lifetime risk from all man-made radionuclides
will not exceed 1 in 1000. Application of the ALARA principle to each
practice or source then leads to further reductions in risks.
The top-down approach also is used in other standards or guidances for
limiting the public's exposure to radiation. These include (1) U.S.
Environmental Protection Agency (EPA) standards for naturally occurring
radionuclides in uranium and thorium mill tailings, (2) EPA guidance on
acceptable levels of radon in homes, (3) a recommendation of the
National Council on Radiation Protection and Measurements (NCRP) on
levels of external background radiation at which remedial actions should
be undertaken, and (4) EPA guidance on appropriate responses to
radiation accidents. In each case, the upper bound on acceptable
lifetime risk is in the range of 1 in 1000 to 5 in 100, and the ALARA
principle is used to reduce risks below these limits.
BOTTOM-UP APPROACH
The framework for regulating routine exposures of the public to chemical
carcinogens and radiation under laws other than the Atomic Energy Act is
what we call a "bottom-up" approach. This approach is essentially the
opposite of the top-down approach described above.
In the bottom-up approach, there is no standard defining an upper bound
on acceptable risk from all carcinogens and sources of exposure. Instead
of limiting public exposures to all sources, standards have been
developed only for specific exposure situations. For each exposure
situation, a lower bound on acceptable risk is established as a goal,
but this goal may be increased, based primarily on cost and technical
feasibility.
The bottom-up approach was first used in the Delaney Clause of the
Federal Food, Drug and Cosmetic Act Food Additives Amendment of 1958.
This law calls for zero risk to the public from carcinogenic food
additives, such as pesticides. However, because zero risk cannot be
achieved at any cost, the EPA usually has permitted carcinogenic food
additives if the lifetime risk is less than 1 in a million--that is, no
more than 1 in a million individuals consuming such food additives over
their lifetimes would be expected to die of cancer resulting from
exposure to these chemicals.
The bottom-up approach next was used in EPA standards for radionuclides
and chemical carcinogens in drinking water developed under the authority
of the Safe Drinking Water Act. These standards specify zero risk from
carcinogens in drinking water as a nonenforceable health goal. However,
because this goal also cannot be achieved at any cost, the standards
then establish legally enforceable limits that must be set as close to
zero risk as possible, taking into account cost and technical
feasibility. Current EPA standards for radionuclides and chemical
carcinogens, which are regulated individually, correspond to upper
bounds on lifetime risk in the range of 1 in 10,000 to 1 in a million.
This approach of defining a range of acceptable risk at these levels has
since been used in two other sets of standards. First, EPA standards for
airborne emissions developed under the authority of the Clean Air Act
include standards for individual carcinogens that are based on lifetime
risks that would not exceed 1 in 10,000 for members of the public
receiving the highest exposures and 1 in a million for the greatest
number of persons in exposed populations.
Second, EPA standards for cleanup of hazardous substances at Superfund
sites developed under the authority of the Comprehensive Environmental
Response, Compensation, and Liability Act (CERCLA) specify that
remediation goals shall consider, among many factors, an upper bound on
lifetime cancer risk in the range of 1 in 10,000 to 1 in a million. In
contrast to the other standards that use the bottom-up approach, the
goal for an upper bound on lifetime risk in this case applies to the
total risk from all carcinogens.
PROPOSAL FOR CONSISTENT REGULATION OF CARCINOGENS
The top-down approach used in regulating exposures to radionuclides
under the authority of the Atomic Energy Act clearly is fundamentally
different from the bottom-up approach used in regulating exposures to
radionuclides and chemical carcinogens under other laws. As a result,
the upper bounds on risks to the public regarded as "acceptable" in the
two cases are clearly inconsistent--upper-bound lifetime risks range
from about 1 in 10 to 1 in 1000 in the former case but from about 1 in
10,000 to 1 in a million in the latter.
This inconsistency is particularly apparent in the case of disposal of
low-level radioactive waste. Currently, acceptable waste disposals under
the authority of the Atomic Energy Act present a maximum lifetime risk
to hypothetical individuals who inadvertently intrude into disposal
facilities of about 1 in 100. However, standards governing cleanup of
old waste disposal sites under CERCLA include, as a goal, an upper bound
on lifetime risk to inadvertent intruders of between 1 in 10,000 and 1
in a million. This considerable difference in acceptable risks for
virtually identical practices seems quite illogical.
We believe that the fundamental inconsistency in current approaches to
regulating exposures of the public to radionuclides and chemical
carcinogens can be reconciled and that a reasonable basis for more
consistent regulation of health risks from exposure to all environmental
carcinogens can be developed. Our proposed regulatory framework, shown
in the figure on p. 25, contains three basic elements:
(1) a de manifestis lifetime risk in the range 1 in 10 to 1 in
1000, which would define an upper bound on acceptable risk from
all carcinogens and sources of exposure and above which
regulatory action to reduce risk would be taken regardless of
cost;
(2) a de minimis lifetime risk in the range 1 in 10,000 to 1 in a
million, which would define risks from any carcinogens and
sources of exposure that are so trivial that regulatory action
to reduce risk is not warranted; and
(3) reduction of risks based on application of the ALARA principle
for lifetime risks above de minimis levels.
The key to our proposal is to recognize that the lifetime risks of 1 in
10,000 to 1 in a million embodied in many standards developed using the
bottom-up approach, in fact, define de minimis (trivial) rather than de
manifestis (mandatory action) levels. In particular, such low risks are
not analogous to the upper bound on acceptable risk implicit in
radiation protection standards, which are based on the top-down
approach.
We would also emphasize that achieving a de minimis risk is not the goal
of ALARA, because the ALARA principle implies a process to be applied to
each exposure situation, not a generally applicable and predetermined
result. It could be reasonable in many situations--for example, after
consideration of costs and benefits--to decide not to reduce risks by
regulatory action at levels well above the proposed de minimis values.
The proposed de manifestis and de minimis risks are given as ranges
rather than single values. This approach would permit taking into
account the size of an exposed population. That is, the higher values
could be used when only a few individuals are at risk, but the lower
values could be used for large populations. The use of ranges also would
permit considerable flexibility in accommodating the kinds of subjective
societal judgments involved in applying the ALARA principle to
particular exposure situations. Thus, absolute uniformity of regulatory
decisions for limiting cancer risks to the public would not be required.
The interpretation of lifetime risks of 1 in 10,000 to 1 in a million as
de minimis, which is a key element of our proposal, is clearly supported
by an analysis of EPA regulatory decisions for chemical carcinogens
before the mid-1980s. This analysis was performed by a group headed by
Curtis Travis of ORNL's Center for Risk Management and Richard Wilson of
Harvard University. As shown in the figure on the facing page, the EPA
always declined to reduce risk by regulatory action when the risk to a
few individuals was below 1 in 10,000 or the average risk in large
populations was below 1 in a million. This is the meaning of de minimis,
and the EPA decision in each case is consistent with our proposed de
minimis levels.
This analysis also indicates a consistency with the other elements of
our proposed regulatory framework. First, the EPA always decided to
reduce risk by regulatory action when the lifetime risk was above 1 in
100 to 1 in 1000, which is consistent with our proposed de manifestis
risk. Second, the EPA took regulatory action in 50% of the cases when
the lifetime risk was between the de manifestis and de minimis levels,
primarily on the basis of expected costs and benefits. This approach is
consistent with our proposed use of the ALARA principle.
Although the EPA regulatory decisions summarized in this figure are
consistent with our proposed regulatory framework, these decisions were
made on a case-by-case basis rather than within the context of an
explicit regulatory framework for all carcinogens and exposure
situations. In contrast to the ad hoc approach previously used by the
EPA, we are advocating that all elements of our proposed framework be
adopted as an explicit set of principles for regulating risks to the
public from all exposures to any carcinogens.
In our discussions of the top-down and bottom-up regulatory approaches,
we indicated that our proposed regulatory framework is consistent with
many standards and guidances for both routine and accidental exposures
to radionuclides and chemical carcinogens. Our proposed framework also
is consistent with exemption (de minimis) levels of radiation exposure
recommended, for example, by the NCRP, which correspond to a lifetime
risk of about 1 in 10,000, and the current action level for
polychlorinated biphenyls (PCBs) in fish, which corresponds to a risk of
1 in 1000.
Therefore, our proposed regulatory framework is consistent with
virtually all current regulatory policies for limiting routine and
accidental exposures of the public to radionuclides and chemical
carcinogens. Again, however, a consistent regulatory framework for all
carcinogens is achieved only if lifetime risks in the range of between
1 in 10,000 and 1 in a million embodied in many standards developed
using the bottom-up approach are interpreted as de minimis.
IMPLEMENTING PROPOSED REGULATORY FRAMEWORK
Our proposed regulatory framework for limiting risks to the public from
all carcinogens and sources of exposure is useful for risk management
but is not concerned with estimation of risks for any exposure
situation. However, we believe that certain important differences in
risk estimation procedures for radionuclides and chemical carcinogens
should be reconciled in implementing our proposed regulatory framework.
The first is an inconsistency in the risk factors that convert exposure
(or dose) to risk. Risk factors for radiation exposure, such as those
recommended by the ICRP, are intended to be best estimates (mean
values). For chemical carcinogens, however, risk factors developed by
the EPA are intended to be upper-bound estimates (95% confidence limits)
and, thus, provide more conservative estimates of risk. In addition,
radiation risk factors take into account cancer risks for all organs of
the body, but risk factors for chemical carcinogens usually consider
only one organ at risk and, thus, ignore risks to other organs.
Second, the primary measure of risk from radiation exposure is fatal
cancers, but the EPA uses cancer incidence as the measure of risk for
chemical carcinogens. For organs in which most cancers are curable, such
as the thyroid gland or skin, risk estimates based on cancer incidence
can be a factor of 10 or more higher than estimates based on fatal
cancers. Radiation risk factors that take into account nonfatal as well
as fatal cancers have been introduced by the ICRP, and similar risk
factors could be developed for chemical carcinogens.
Finally, in assessing radiation exposures, the intent usually has been
to provide best estimates of dose using reasonable assumptions for
likely exposure scenarios. However, risk assessment procedures
prescribed by the EPA for use at Superfund sites, for example, often
emphasize unreasonably pessimistic assumptions. Thus, the resulting risk
estimates may greatly exceed values that reasonably could be
experienced.
CONCLUSIONS
We believe that more consistent regulation of risks to the public from
exposure to radionuclides and chemical carcinogens along the lines
proposed here would have two obvious benefits. First, it would encourage
consideration of risks from any carcinogen and source of exposure in the
context of the total cancer risk from all sources. In the past, the EPA
has undertaken regulatory actions in a rather piecemeal fashion,
particularly for chemical carcinogens, primarily because of inconsistent
requirements in the many laws under which the EPA operates.
Second, the proposed range of de manifestis risks is consistent with
lifetime risks from naturally occurring carcinogens. Natural background
risks average about 1 in 100 for radionuclides and at least 1 in 100 for
chemical carcinogens. Therefore, the proposed range of de minimis risks
would ensure that cancer risks much less than the largely unavoidable
background risks do not receive unwarranted attention.
Can we hope that a more consistent regulatory framework for all
environmental carcinogens eventually will be implemented by the EPA? Two
encouraging developments suggest that the inconsistency between the
top-down and bottom-up regulatory approaches and the need to reconcile
them have been recognized. First, the EPA office that administers the
Superfund program recently indicated that site remediation need not be
undertaken if the maximum lifetime risk to individuals is below 1 in
10,000 and that higher risks would be acceptable when risk reduction is
not feasible. Such a policy should help overcome the widely held and
unreasonable view that risks above 1 in 10,000 are unacceptable (i.e.,
intolerable).
Second, the Radiation Advisory Committee of the EPA's Science Advisory
Board recently addressed former EPA Administrator William Reilly on the
need to reconcile the inconsistent approaches to risk reduction
currently used for radionuclides and chemical carcinogens. The
committee's primary concern was the evident inconsistency between
established guidance on acceptable levels of radon in homes, which
corresponds to a lifetime risk greater than 1 in 100, and a proposed
standard for radon in drinking water, which corresponds to a risk at
least a factor of 100 lower. However, the committee also urged Reilly to
consider more consistent approaches to regulating risks from all
environmental carcinogens.
The use of inconsistent regulatory policies for environmental
carcinogens could unnecessarily increase the cost of complying with
regulations, particularly for cleaning up hazardous waste sites under
the Superfund program. Thus, the public clearly has an important stake
in efforts to promote more consistent regulation of environmental
carcinogens to ensure that money is spent wisely to reduce health
risks.Biographical SketchesDavid C. Kocher has served as an
environmental health physicist in ORNL's Health and Safety Research
Division since 1976. He is also a faculty affiliate in the Department of
Radiological Health Sciences at Colorado State University and a frequent
lecturer on environmental dose assessments and radioactive waste
disposal. He received his Ph.D. degree in experimental nuclear physics
from the University of Wisconsin. Following a postdoctoral appointment
at the University of Birmingham in the United Kingdom, he joined ORNL's
Physics Division in 1971. In his health physics research at ORNL, he has
developed models for estimating radiation doses to the public from
radionuclides in the environment and has been concerned with the
scientific basis for environmental regulations. F. Owen Hoffman, former
ORNL ecologist, was recently named president and director of SENES Oak
Ridge, Inc., Center for Risk Analysis. After employment with the U.S.
National Park Service and the Institute for Reactor Safety in Cologne,
Germany, he worked for ORNL's Health and Safety Research Division and
then the Environmental Sciences Division from 1976 to 1992. He serves as
chief scientist to the International Atomic Energy Agency in Vienna on
the validation and evaluation of radiological assessment models. He has
visited the former Soviet Union to participate in a joint investigation
of the environmental behavior of radionuclides from the Chernobyl
accident. Recently, he was appointed by the governor of Colorado to
serve on the Colorado Department of Health's Advisory Panel for the
assessment of contaminants at Rocky Flats; by Louis Sullivan, then
secretary of the U.S. Department of Human and Health Services, to serve
on the Centers for Disease Control advisory panel studying deaths from
thyroid disease near Hanford, Washington; and by the Tennessee
Department of Health to serve on the Health Advisory Steering Panel on
radiation dose reconstruction for the Oak Ridge Reservation.
David C. Kocher and F. Owen Hoffman
(keywords: carcinogens, wastemanagement, environmental regulation))
------------------------------------------------------------------------
Please send us your comments.
Date Posted: 1/11/94 (ktb)