Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
InI;roduction Through Congress the American public has granted author- ity to Federal administrative agencies to restrict private actions, such as the production and use of chemicals, when this is deemed necessary to protect the health of the public. The 1970 s are notable for the large number of new federal regulatory laws that are applicable to the environment, both in the workplace and in the community. These laws reflect a dramatic and relatively rapid shift in public priorities toward the protection of health. Concurrently with shifts in social priorities, advances in science have contributed to policy problems, for the advances have revealed the extent of the environmental health problem. Some earlier regulatory programs had addressed exposure to toxic chemicals, but they were directed mainly at the risk of poisoning and other acute effects. Much policy-mak~ng related to such effects involved routine, short-term, acute animal studies to establish ~no-observed-effect" doses and then the straightforward calculation of allowable hen exposure based on the application of safety factors to relatively uncomplicated scientific findings. Such an approach reflected little recognition of problems that might be associated with smaller exposures. Cancer, birth defects, and other conditions were seldom seen as preventable by government intervention. Only in the last 15 years has the potential extent of the linkage between such condz- t ions and toxic substances been revealed. The often- c~ted estimate that a large fraction of all cancers may be attributed to human exposure to toxic agents (including smoking, diet, lifestyle, and occupation) originated f airly recently (Boyl and, 1969 ; }~gginson, 1969), and it 9
lo was not until the 1970s that regulatory agencies focused their attention on cancer and other chronic health risks. Scientific advances entered the picture in a second way. The technology that has made it possible to detect relations between particular agents and cancer or other chronic effects has evolved rapidly from the days when exposure through skin~pa~nting and subcutaneous inj action care re~'ea on in animal tests of carcinogenicity. }n- creasingly, ep~dem~otog~c investigations have earner confirmed the findings of animal experiments of provided evidence that linked exposures to particular chemicals to particular chronic health effects. The introduction of reliable testing methods resulted in broader government testing requirements and, steadily, the discovery of more and more suspect chemicals--many of them in common use~- that demanded agency attention. The techniques are still developing, and we are still looking for better ways to design and interpret annoy bioassay experiments. The increase in newly suspect chemicals was accompanied by the development of instruments and procedures That permitted the detection of chemicals at lower and lower concentrations. Even if the number of suspect chemicals had not increased dramatically, these sensitive detection methods would have revealed the presence of such chemicals in concentrations that earlier methods would have missed. Combined with all those changes were the development and refinement of analytic methods of estimating the degree of human risk on the basis of data from hewn studies and ~ ~ _ ant mad expert menus . Public policies are not immediately adaptable to rapid changes in social priorities and scientific advances. Many of the fundamental difficulties of regulatory risk assessment result from attempts to bend old laws and policies to fit newly perceived risks. For instance: · A regulatory framework based on she traditional approach involving norobserved-e~fect doses and safety factors is now being applied to health effects for which a no-effect dose cannot be demonstrated, except at zero exposure. · Regulatory laws and programs designed for the elimination of what was understood to be the very rare event of chronic hazard now operate in the presence of the recognition that many agents are suspect. · Agencies must evaluate hundreds of chemicals on which no data related to human risk are available and on
11 which few animal tests were required and many other chemicals that were tested with methods that do not meet modern standards. . Laws were written and programs designed before current quantitative methods for estimating human risks on the basis of data from animal studies were developed. DIFFICULTIES IN DECISIONWRING Agency decisions regarding potential carcinogens and similar hazards are commonly beset by two types of dif- ficulties: inherent limitations on the power of analysis and practical constraints imposed by external pressures. Several such factors are particularly relevant to the consideration of scientific aspects of risk assessment. INHERENT LIMITATIONS Uncertaintv The dominant analytic difficulty is pervasive uncertainty. Risk assessment draws extensively on science, and a strong scientific basis has developed for linking exposure to chemicals to chronic health effects. However, data may be incomplete, and there is often great uncertainty in estimates of the types, probability, and magnitude of health effects associated with a chemical agent, of the economic effects of a proposed regulatory action, and of the extent of current and possible future human exposures. These problems have no Immediate solutions, given the many gaps in our understanding of the causal mechanisms of carcinogenes~s and other health effects and in our ability to ascertain the nature or extent of the effects associated with specif ic exposures. Because our knowledge is limited, conclusive direct evidence of a threat to human health is rare. Fewer than 30 agents are definitely linked with cancer in humans (Tomatis et al., 1978); in contrast, some 1,500 substances are reportedly carcino- genic in animal tests, although they include substances tested in studies of questionable experimental design. We know even less about most chemicals; only about 7,000 of the over 5,00O,000 known substances have ever been tested for carcinogenicity (Maugh, 3978)--a small fraction of those theoretically under regulatory Jurisdiction. We
12 know still cancer. about chronic health effects other than Ethical considerations prevent deliberate human experi- mentation with potentially dangerous chemicals, and the length of the latent period for cancer and some other effects greatly complicates epidemiologic studies of uncontrolled human exposures. Animal models must be used to investigate whether exposure to a chemical is related to the incidence of health effects, and the results must be extrapolated to humans. To make judgments amid such uncertainty, r Ask assessors must rely on a serf es oF assumptions . Limited Analytic Resources The number of chemicals in the jurisdiction of federal regulatory agencies is enormous. For example, of the roughly 5, 000, 0 00 known chemicals, more than 70,000 are in commercial use (Fishbe~n, 1980). The Environmental Protection Agency's Chemical Activities Status Report lists about 3, 500 chemicals as being under some sort of active consideration in the Agency's various regulatory programs. Similarly, the Food and Drug Administration's food program must cope with over 2,000 food-related chemi- cals (900 flavors, 700 items listed as generally recog- nized as safe," 350 food additives, 175 animal drugs, and 60 color additives) and an additional 12,000 indirect additives (Flamm, 1981). The many problem chemicals in an agency jurisdiction compete for attention of analysts and decision-makers. If an agency is considering new action on many substances at once, its scientific staff is stretched thin. Most agencies do not have the analytic resources to do a thorough risk assessment for priority-setting anc..~st rely on less formal methods to ensure that the highest- risk chemicals are examined first. Complexity For most chemical agents that might be subject to regular tion, a great variety of factors must be assessed, includ- ing potential toxicity, extent of human exposure, effec- tiveness of technologies to reduce exposure, the nature of possible substitute chemicals, effects on and interests of various population groups, and economic effects of
13 regulatory alternatives. Decisionrmake~s In a regulatory agency may encounter a large amount of highly technical information as they work toward their decisions; many scientific disciplines and technical fields are usually involved. An agency would like to have simple rules and analytic procedures to ensure consistency and competence in its decision-making, but, in the face of scientific uncertainty, such simplicity is cliff Cult to achieve without an inadvertent loss of crucial scientific insight from the decision process. EXqE:RNAL P~:SSURES l- Public Concern with Health Protection When the risk involves a serious disease, such as cancer, or birth defects, feelings are likely to run high, par- t~cularly if the groups exposed to a chemical are mob~l- ized to express themselves in an agency's deliberations. Such groups insist that regulatory action need not await conclusive evidence of cause and effect and need not be based exclusively on the most scientifically advanced testing methods. Visible Economic Interests Although it is rarely known which individuals are likely to be saved from adverse health effects through a regu- lation that reduces exposure to a particular chemical, those who bear the economic costs of such restrictions can identify themselves without any difficulty. These parties can provide relatively concrete projections of a prospective regulation's inflationary influence, effect on employment, and other immediate economic effects, and such consequences may be substantial. They may question the wisdom of balancing concrete evidence of economic damage against evidence of health protection that depends on a complex series of assumptions derived from sparse and indirect data. Congressional Action In fulfilling its role as the legislative voice of popular concerns, Congress can act in ways that influence decision
~4 processes. It can dictate the factors to be included In and excluded from decision making t the Delaney clause is an example), and it can pass special legislation to prep empt agency discretion, as it did in acting to prevent the removal of saccharin from the market. PROPOSED REFORMS Under these conditions' it would perhaps be surprising if calls for major reform were not heard. Some have sought to improve the techniques that the government uses to analyze and evaluate risks; for example, the House of Representatives In 1982 passed H.R. 6159 (commonly known as the fritter bills), to establish a government-wide program of research and demonstration projects on quanta titative and comparative risk analysis. Much of the recent controversy is general; it reflects the conflict in values between different groups in soc:- ety, particularly with regard to the relative importance of economic factors and health protection in the formula- tion of regulatory decisions. Different groups w~11 inevitably disagree about the degree of risk (if any) that is defined as acceptable in a particular case. However, some criticisms directly address the risk assessment component of the overall decisionrmaking process. Some critics question whether current practices adequately safeguard the quality of the scientific interpretations needed for risk assessment. With a scientific base that is still evolving, with large uncertainties to be addressed in each decision, and with the presence of great external pressures, some see a danger that ache scientif ic interpretations i n risk aSses=~'ents will be distorted by policy considerations, and they seek new institutional safeguards against such distortion. Among the institutional reforms suggested, two major _;: an" Who focus of this refit: reorganization Acid ~ =~__ ~ ~ _ ~~ _ to ensure that r isk assessments are protected f ram inappropriate policy influences and development and use of uniform guidelines for carrying out risk assessments. Same argue that scientific quality, consistency, and distinction between scientific judgment and policy 3udg- ment can be Improved through the use of explicit guide- lines for agency risk assessments. Such guidelines would specify methods for interpreting scientific data and would seek to limit analysts who confront data gaps or
IS extrapolation questions to methods that are consistent with the best current scientif ic judgment. Analysts following the guidelines would find it easier to describe systematically and explicitly the methods that are incor- porated in their risk assessments. Several other recent proposals call for major restruc- turing of federal processes to separate the risk assess- ment function organizationally from dec~szon-making. The objectives would be to permit analysts to work indepen- dently of policy pressures and to foster consistency of risk assessments. Various approaches have been suggested, including creation of a single body outside the government for the performance or review of risk assessments, crea- t~on of a single government unit to conduct risk assess- ments for the entire government, and creation of separate risk assessment units in particular programs or agencies and systematic review of assessments by independent scientific advisory groups. The STODY This report responds to a congressional request to examine the merits of the two major types of Reform proposal. It is the final report of the National Research Council's Committee on the Institutional Means for Assessment of Risks to Public Health. Chapter ~ describes the structure of risk assessment, the role of science in the assessment process, and current federal uses of risk assessment. Chapter IT examines the feasibility and desirability of the development and use of uniform guidelines. Chapter III reviews various organizational arrangements for risk assessment. The Committee's overall conclusions and recommendations appear in Chapter IV. REFERENCES _ . _ Boyland, E. 1969. The correlation of experimental carcinogenesis and cancer in man. Prog. Exp. Thor Res. 77: 222 - 234 . F i shbein, L . 1980. Potential industrial carcinogenic and mutagenic alkylating agents, pp. 329-363. In D. B. Walters, ed. Safe Bandling of Chemical Carcinogens, Mutagens, Teratogens, and Highly Toxic Substances, vol. I. Ann Arbor, Mich.: Ann Arbor Science.
16 Flame, W. G. October 13, 1981. Remarks to the Committee on the Institutional Means for Assessment of Risks to Public Health. 1969. Higginson, J. 1969. Present trends in cancer epidemi- ology. In Proc. Can. Cancer Conf. 8:40-75. Maugh, T. 1978. Who chooses chemicals for test ing? Science 201:1200. Tomatis, L., C. Ag the, B. Bartsch, 3. Buff, R. Montesano, R. Saracci, E. Walker, and J. W~lbourn. 1978. Evaluation of the carcinogen~c~ty of chemicals: a review of the monograph program of I.A.R.C. Cancer Res. 38:877.