National Academies Press: OpenBook

Cost of Environmental-Related Health Effects: A Plan for Continuing Study (1981)

Chapter: Chapter 3: Information Needed to Assess and Quantify Health Effects

« Previous: Chapter 2: Data Related to Environmental Hazards and Human Exposure
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 49
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 50
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 51
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 52
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 53
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 54
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 55
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 56
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 57
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 58
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 59
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 60
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 61
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 62
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 63
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 64
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 65
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 66
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 67
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 68
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 69
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 70
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 71
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 72
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 73
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 74
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 75
Suggested Citation:"Chapter 3: Information Needed to Assess and Quantify Health Effects." Institute of Medicine. 1981. Cost of Environmental-Related Health Effects: A Plan for Continuing Study. Washington, DC: The National Academies Press. doi: 10.17226/812.
×
Page 76

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.

CHAPTER 3 INFORMATION NEEDED TO ASSESS AND QUANTIFY HEALTH EFFECTS Repeated attempts in recent years to assess the health impacts of environmental factors have been hampered by the fragmentary state of present knowledge. Public Law 95-623 specifically lists cancer, birth defects, genetic damage, emphysema, asthma, bronchitis and other respiratory diseases, heart disease, stroke, and mental illness and impairment as examples of health problems to be considered by the ongoing study. However, environmental agents may also damage the skin, kidneys, immune system, and other organs of the body, and the committee believes that these and other effects should not be neglected. Appendix C contains tables of known and suspected environment-related health effects. This chapter continues the discussion of information needed to carry out the ongoing study, with particular reference to the methods used to ascertain the health effects of environmental hazards (Boxes 4 and 4a in Figure l-l).* Viewed in modern perspective, the development of virtually all diseases may be presumed to involve an interaction between genetic and environmental determinants.] The contribution of a given environmental risk factor can be expected to vary, depending on the conditions under which it is encountered, the presence of other factors in the environment that may modify its effects, and the susceptibility of the exposed population. Although adverse effects of various environmental factors have been well documented when such effects have occurred promptly in a high proportion of exposed individuals, effects have been more difficult to document when they have been delayed in their appearance, have occurred in only a small proportion of exposed individuals, or have resulted from the combined effects of two or more factors.2 Yet, it is the effects that occur under the latter conditions that are of greatest concern in assessing the impact of the environment on the health of the population as a whole. . Although this committee suggests that the ongoing study focus on the health effects resulting from involuntary exposure to physical and chemical pollutants, it will often be necessary to consider additional factors such as life-style factors and socioeconomic status. -49-

Three types of studies provide evidence about whether a substance or exposure may lead to health problems o short-term experimental in vitro tests* and animal studies o clinical studies, particularly studies using human volunteers o epidemiologic studies, which provide data on the distribution and determinants of disease and illness in human populations. Each of these methods has strengths and weaknesses, and there is a great need to integrate data obtained by the three methods.3 Short-term tests and animal studies can predict potential toxicity before actual human exposure occurs. Short-term tests primarily detect substances that interact with genetic material,4~8~9 but animal studies can detect a wide range of toxic effects, including behavioral effects and damage to liver, skin, the nervous system, and other organs. With animals, acute effects, as well as those with long latent periods, can be studied. Dose/response curves can be derived, and the effects of substances on target organs analyzed. However, difficulties arise in applying data from short-term or animal studies to people. In some cases, evidence of health problems resulting from exposure to hazards noted by clinicians can provide a link between animal and short-term experimental studies on one hand, and epidemiologic studies on the other. Controlled experiments with toxic agents on human beings are constrained by practicality and ethics, so that they are rarely used today and usually involve relatively few (1-10) volunteers exposed to agents whose physiologic effects are easily measured and reversible, such as sulfur oxides.5 Epidemiologic studies provide statistical associations between the factors of concern and health effects in the population. They provide information on real people in a real environment and, therefore, are particularly useful for purposes of the ongoing study. The necessary use of observational investigation in most human epidemiologic studies means that such studies generally are not useful for predicting health effects before chemical or physical agents are introduced. There are major difficulties in detecting or estimating health effects for large populations at low levels of exposure. Estimation of risks at low doses requires either large population samples or extrapolation of effects at high doses to effects at low doses, with all the uncertainty such extrapolation procedures entail. The *Short-term tests refer to a group of tests, using bacteria or other biological systems, that detect substances that are potentially . mutagenic or carcinogenic. —SO—

effects of radiation on human beings provide one of the better examples of an area where data on both exposure and health effects exist. Even in this example, however, controversy continues about estimating the health effects of low levels of radiation because of the need to extrapolate from results seen at higher doses.2 Epidemiologic studies have become more informative as researchers have learned to control for confounding factors, and to do intricate statistical analyses. Some difficulties can be overcome by study design and by collecting data on particular populations, such as workers or susceptible groups, and by taking advantage of accidental events that lead to high levels of exposure. RISK EXTRAPOLATION FROM NON-HUMAN DATA Most environmental hazards have been ;denti fled after they caused human illness or disease. However, recent efforts are directed at predicting substances likely to cause health problems, and at minimizing or preventing human exposure. Attention is directed particularly to the effects of chemicals, because expansion of chemical technology since World War II has vastly increased both the variety and volume of chemicals used in the United States. An inventory by the Environmental Protection Agency (EPA) lists more than 55,000 chemical substances that are used commercially and are subject to regulation by the Toxic Substances Control Act (TSCA).6 The inventory excludes chemical mixtures, foods, drugs, cosmetics, pesticides, and other substances regulated under authorities other than TSCA. With rare exceptions, these materials have not been demonstrated to be either safe or unsafe to humans. Short-term Tests An array of short-term tests has been developed to provide relatively rapid and inexpensive means for screening compounds for mutagenic or related effects.4~8~9, Current short-term tests involve a variety of biologic systems, including bacteria, yeast , fruit flies, and cultured mammalian cells. Conducting short-term tests on a compound takes a few days to a few months and costs from a few hundred to a few thousand dollars . Positive results in these tests are highly correlated with the ability of certain classes of subs Lances to cause cancer in animal tes ts, 4 ~ 7 ~ ~ ~ 9 a [though some materials that are carcinogenic. by other criteria do not give posi t ive resul ts in many of these tests . The methods have been reviewed, ~ and an international effort that recently compared results obtained for many of the standard tests in different laboratories concluded that further development and validation are needed.l° -51-

Although short-term tests are used primarily for detecting mutagenic and other effects that are assumed to be related to carcinogenesis, it would be useful if there were quick tests for other types of effects--such effects could include problems with reproductive or nervous system, liver function, or lung function. Researchers now are attempting to develop such tests. Animal Tests Because there are basic similarities between human beings and other animals in the way their cells and tissues respond to toxic or hazardous agents, animal experiments can be helpful in evaluating the human health risks from many substances. For a number of compounds,12 such as vinyl chloride, aflatoxin, and diethylsti lbesterol (DES) ~ harmful effects were detected in animal studies before such effects were recognized in human beings. Animal tests are widely used for detecting acute effects of substances . More attention i s needed to develop and use anima1 studies to detect diseases with long latent periods, such as cancer, respiratory diseases, heart disease, and neurological problems. Requirements for adequate numbers of animals, appropriate controls, and sufficient length of time to allow the expression of health effects that have long latent periods make such experiments expensive. Even using small rodents, testing two species for carcinogenicity ordinarily requires more than SOD animals, exceeds $500,000 in cost, and takes up to three years.13 Pisk Extrapolation Tests that do not involve human beings raise questions of whether results from the test system apply to human beings, and, if so, how to express the relationships quantitatively. For example, although the reactivity of DNA is similar whether situated in a test bacterium or a human cell, there are large differences in the environments traversed by a chemical on its way to the target DNA and in the organization, regulation, and repair of DNA in the two cases. It is not possible at present to use short-term test results to estimate human risk quantitatively. In animal tests for carcinogens, animals receive high doses of the substance under study, so that a reasonable percentage of them will develop tumors if the substance is carcinogenic. Risk estimation depends on dose/response relationships, and many attempts have been made to develop models to extrapolate results from high doses in animals to the low doses to which people are more likely to be exposed .2 ,14-17 -52-

Risk extrapolation estimates using various models give similar results at high doses where measurements are available, and give very different results at low doses. In one massive experiment, called the ED01 study, more than 24,000 mice received varying doses of a carcinogen so that low-dose effects would be detectable and dose/response models could be tested.15 The number of cancers observed were consistent with several different models and demonstrated the difficulties in risk extrapolation. At present there is not a good theoretical basis for extrapolating results of animal studies to estimate human risk.18 One of the systematic efforts to make quantitative comparisons of the carcinogenesis of chemicals in mice, rats, and human beings is that of Crouch and Wilson.l9 Their method has shown a high correlation for most of the compounds tested and begins to offer a basis of rational estimates of human risk from animal data. However, further study of compounds that show idiosyncratic reactions may lead to discovery of useful comprehensive principles for estimating human risk from experimental species. Pisk extrapolation is at a stage of conceptual development such that a wide range of data scattered among many different sources needs to be integrated. Quantitative judgments, although necessary, are particularly difficult. Much medical research is needed--from fundamental molecular biology to more immediately applicable medical and public health practices--in order to develop a theoretical basis for predicting possible human health effects from studies on non-human test systems. Improved understanding of the ways in which environmental agents affect different biological systems should provide a basis for developing less costly and more rapid tests to estimate human risk. CLINICAL STUDIES Clinical studies may provide data to help detect and quantify the health effects of environmental hazards in at least two ways. First, the medical care system, either by observations of individuals or by analyses of records, can detect the occurrence of unusual health problems or identify cases of disease associated with exposure to an environmental hazard. This will be discussed further in the following section on epidemiology. Second, in some circumstances, clinical studies can be carried out to evaluate health effects after exposure to a pollutant under rigorously controlled conditions.5 However, these studies are sharply limited by ethical concerns; exposures must be limited to those that cause only temporary effects and do not cause volunteers undue discomfort or pain. Clinical observations, whether they are case reports associated with exposures or the result of data collected in controlled studies, can help relate data obtained in animal studies to those obtained in -53-

epidemiologic studies. is similar to an animal from animal studies. Clinical documentation of human disease that disease permits more secure generalization EPIDEMIOLOGIC STUDIES The relevance of epidemiologic principles and methods to the ongoing study is evident from the common definition of epidemiology--the study of the distribution and determinants of disease frequency in human populations. The unit of study--human populations--is emphasized. In principle, given the congressional charge in P.L. 95-623, all the health consequences of exposures to all the hazards deriving from the man-made environment should be assessed in all residents of the United States, and monitored over time. Although this is an impossible task, it is desirable to retain the concept, in order to discuss the types of epidemiologic studies that are or should be carried out and their limitations. The major strength of epidemiology is that it provides a direct measure of risk in human beings. However, the technique has limitations, especially for conditions that have long latent periods and occur after low levels of exposure to hazardous agents. Studies in these instances require large populations, adequate exposure data, and careful analysis of confounding factors.* The necessary use of the observational mode of investigation in most human environmental epidemiologic studies imposes limitations on the causal inferences derivable directly from relationships between exposures and diseases. Types of Studies Epidemiologists often classify studies according to the purpose of the study and the research designs employed to achieve the purpose. Descriptive studies can be used to monitor and to document the burden of illness and disease from known hazards, and to generate hypotheses for future testing in situations where the adverse health consequences of substances found in the environment are not yet known. Analytic studies can be used to test specific hypotheses about environment-health relationships. The major types MA confounding factor is a factor that contributes to a disease incidence, and to which exposure frequently occurs under the same conditions as exposure to the substance whose effects are being studied. An incorrect estimate of the risk attributable to the substance under study will be made if confounding factors are present but ignored. For example, if respiratory illness is attributed to the air pollutant sulfur dioxide but is in reality due primarily to asbestos exposure that occurred in association with sulfur dioxide, then asbestos is a confounding factor. -54-

of study designs used for these purposes are described below.* With few exceptions, the greater the analytic power of the design, the more complex the study and the more resources will be needed to carry it out. (1) Ecologic studies look at patterns of morbidity and mortality in aggregates or groups of individuals in relation to information collected on an environmental. characteristic. There is no direct measure of individual exposure or other individual characteristics. An example is a comparison of cancer in counties of the United States according to the presence of selected industries. Because the individuals with cancer in such a study are not classifiable according to how near they live to the industry or whether they work in that industry, no firm association between cancer and the industry can be made. The hypothesized relationship needs further testing. (2) Case-series studies start with individual cases of a disease, often identified by an alert physician who suspects or discovers a common factor associated with the cases.22~23 The discovery that vinyl chloride is associated with angiosarcoma is an example of a case-series study. A surgeon noted that three patients with hepatic angiosarcoma worked in the same factory and were exposed to vinyl chloride. Vinyl chloride has since been shown to cause that rare cancer. (3) Cross-sectional studies evaluate differences in health status at a specified time among individuals in two or more groups exposed to different levels of the environmental factor or factors in question. In the Community Health and Environmental Surveillance System (CHESS) and in the Six City Study, a group of communities was selected to represent an exposure gradient for designated pollutants so as to use the cross-sectional design as one of the study strategies.24-26 The surveys on health status of the U.S. population carried out by the National Center for Health Statistics, for example, the 1980 National Natality and National Fetal Mortality Surveys, are good examples of cross-sectional epidemiologic studies. These classifications are flexible. In some cases, a complex study might fit into more than one category. In addition, these studies are described in terms of exposure to hazardous substances, but the studies could look for relationships to life-style factors well. -55-

(4) Case-control studies begin with two groups of individuals that differ according to whether or not they have a specific disease or illness. The individuals with the disease or illness (cases) are compared with those who are free of the disease (controls) for past exposures to factors that might be associated with the illness or disease. This type of study can estimate the relative risk of disease associated with exposure, but cannot give information on the absolute risk for a population. As an example of a case-control study, Newhouse used this design to show that inhalation of asbestos particles increased the chances of contracting - mesothelioma, a rare cancer of the lining of the chest or abdomen.27 (5) Cohort studies evaluate the differences in health status emerging over time among specified groups. The groups, or individuals within the groups, may be classified according to the level of past exposures to specific agents and by other characteristics, such as age, occupation, or residence. Health status and further exposures are followed over time. This study design is more powerful than a case-control or cross-sectional study.- It provides information on the absolute risk associated with exposure, as well as the relative risk among the different exposure groups. Cohort studies require long-term commitment of resources. One example of a cohort study is the Atomic Bomb Casualty Commission's study of the Japanese populations exposed to radiation in Hiroshima and Nagasaki, a study that has continued for more than 35 years.28 a less expensive technique that can be used to establish a cohort involves linking data that have already been collected on individuals for other purposes. For example, in England a 1 percent sample of the population is followed longitudinally by linking various records of individuals in order to study occupational health.29~30 Establishing Causal Relationships If~associations are demonstrated by rigorous unbiased epidemiologic studies, then the problem is to demonstrate a cause and effect relationship between a particular entity in the environment and a particular health effect. Epidemiologists establish causality, with greater or lesser certainty, by fulfilling as many as possible of the following rules of evidence for establishing causality (1) Strength of association The stronger the association between a definable environmental agent and the presence of an -56-

observable adverse health effect, the more likely it is that the relationship is causal and not attributable to known or unknown confounding factors. (2) Dose/response gradient The finding of increased incidence of disease with increasing exposure to the substance in question suggests a biologic effect. When the environmental agent or substance is removed, the effect should diminish or disappear. For the purposes of the ongoing study, it would be important to quantify the dose/response gradient. (3) Confirmation of the study Confirmation of the study's finding by other researchers in other populations under different circumstances decreases the chance the association is an artifact. (4) Biological plausibility Are the effects in humans consistent with laboratory findings? (5) Temporal sequence Because cause must precede effect, studies that record exposure and disease status without regard to which came first may be less reliable for inferring causal relationships than studies that have a clear time frame and collect data over a long period of time. (5) Consistency with animal experimentation When the results of animal experiments agree with those found in epidemiologic studies, the likelihood that there is a cause and effect relationship is increased. (7) Specificity of response The finding that exposure to a particular substance always results in a particular disease would be strong evidence of a causal relationship (the one-cause, one-disease concept). However, this criterion is rarely available to epidemiologists studying environmental health effects and, in addition, must be used with care. Almost all diseases of interest have multiple causes, and specific exposures generally are accompanied by a multiplicity of health consequences. ~ slightly different but similar approach to that above for establishing causality for environment-related health effects has been proposed by Hackney and Linn.31 They updated Koch's postulates to apply to modern problems of environmentally caused health effects. Even if ~ causal relationship is established, the fact that many health problems have multiple causes must be considered in projecting the results of epidemiologic studies to anticipated costs and savings in reducing these problems with control measures. Not only are the responses of individuals conditioned by genetic fac tars, but antecedent life experiences, including nutrition, exposure to —5 7—

infectious agents, and exposure to physical and chemical environmental agents other than the one under inquiry, also will modify the biologic response to the causal agent. Analysis of Data from Varied Sources Large bodies of data have been collected on various characteristics of the environment and on the health status of individuals. These data, although collected for other purposes, can be used in ecologic study designs to monitor for known hazards and to generate new hypotheses regarding presently unknown hazard-disease causal relationships for further testing. The strongest kind of evidence in establishing cause and effect relationships comes from data on individuals in cohort or case-control studies. Such studies are often done by linking records of various kinds, such as birth records, death records, social security and health records, to obtain information from different sources. One very successful example of linking data on individuals to study occupational health is the 1 percent sample studies in Britain which define the cohort to be followed over time.29~30 In the United States, linking data on individuals in one data set with information on the same individual in another data set would be greatly facilitated by the use of a "person number" or a unique personal identifier. For example, it would be possible to link health, medical, and census records to seek new insights into factors associated with illness. People could be followed prospectively, or past records could be retrieved. Such personal identifiers are used in other countries; NCHS discussed their use recently. ~ recent study by NCHS related to locating, assessing, and treating individuals exposed to hazardous substances indicated the difficulty of tracking people in the United States, partially because existing statutes--including the 1974 Privacy Act (U.S.C.552a)--protect individuals against undue invasion of privacy by the federal government.32 Because the issues surrounding the use of a personal identifier in the United States-are complex, the planning committee observed that the subject deserves further study. When possible, the ongoing study should use existing data systems to obtain the necessary information. However, as noted earlier, ideal data systems are nonexistent, and the problem is one of adapting available information for the immediate purpose. Longitudinal studies, that is, studies carried out over extended time periods, are probably necessary for answering certain research questions. Notable examples of longitudinal studies of special populations include the National Academy of Sciences/Atomic Bomb Casualty Commission studies of persons exposed to radiation from atomic bomb explosions. These studies documented early effects on fetuses of exposed women, and decades later, excess rates of various cancers.28 The Framingham Heart Study, supported by the National -58-

Heart, Lung, and Blood Institute, has provided much useful inflation about risk factors for heart attacks.33,34 Although the power of well-designed longi tud i nal studies is great, they require long-term commitment of personnel and resources. See Appendix E for approximate costs associated with some epidemiologic studies. Surveillance In addition to being able to use the various epidemiologic research designs for deriving causal relations, epidemiology can also be used for purposes of surveillance. Surveillance serves two purposes. The first of these is to monitor for known hazards and their health consequences, for example, to monitor the workplace. The second is to generate hypotheses that can be tested with further research. Complete Population and Integrated Environmental Surveillance The identification of all exposures and all disease episodes among all individuals would provide the totality of observational information available to identify environmental hazard-disease relationships. Although there are no instances of total surveillance, attempts have been made to implement portions of a general system through the use of population registries, disease registries, and environmental monitoring systems. Population registries have a wide range of objectives, from identifying all citizens of a country--their residences, family relationships and vital statistics--primarily for legal purposes, but also useful for health studies,35 to identifying an exposure cohort, such as the Japanese survivors of the atomic bombings, for follow-up to determine long-term health effects.28 Disease registries have been developed for objectives besides etiologic research; for instance, cancer registries have been developed to improve diagnostic and therapeutic practices, professional education, and assessment of treatment efficacy. When the community catchment areas and population at risk estimates are defined and case ascertainment and reporting are relatively complete and of good quality, the registries can serve several purposes. The cancer registry in Connecticut, for example, has been used to estimate-the incidence-of selected cancers and their relation to potentially hazardous environmental exposures, including those associated with industry and occupation.35 When the risk of a malignancy is known to be markedly increased upon exposure to a specific environmental agent--such as risk of mesothelioma and exposure to asbestos, 27 angiosarcoma and vinyl chloride,22 and bladder cancer and beta-naphthylamine36--combined exposure and disease registries may be used to identify populations previously not -59-

known to be exposed and to measure the success of efforts to remove the environmental hazards and to decrease env~ronment-related disease over time. Environmental monitoring systems are numerous,37 have had a wide range of objectives, and have been applied to many target areas; there are, however, fewer examples of systematic efforts to measure specific exposures for defined populations in whom disease or illness is also ascertained. As two examples, CHESS and the Six City Study have measured air quality of communities and selected respiratory symptoms in resident children and adults.24-26 In contrast to the specific environmental measurements, community orientation, and concurrent ascertainment of acute symptoms of these two studies, environmental epidemiologic studies often have to rely on estimations of exposures, sketchy work histories, and poorly documented subsequent disease that may not manifest itself for decades after exposure. Survey Samples There have been attempts to achieve more limited objectives than the goals of complete population and integrated environmental surveillance, by use of information derived from samples of environmental measurements and targeted surveys of subsets of the population. The surveys of the National Health Survey (~s),38 including periodic household interviews (National Health Interview Survey), examinations (National Health and Nutrition Examination Survey), utilization of health facilities (National Hospital Discharge Survey), and physician/patient encounters in ambulatory settings (National Ambulatory Medical Care Survey), and surveys by the National Institute for Occupational Safety and Health (NIOSH), including probability samples of industrial facilities for environmental evaluations, are examples of efforts to obtain estimates of national experience. In addition to the economy of scale, these surveys have the advantages of improved quality through expert design and centralized control. They have limitations for the purposes of the ongoing study because of the usual absence of follow-up data for individuals in these surveys and the usual absence of linkage of exposure and health measurements. The NIOSH survey omitted large groups of workers and collected no information on dose/response relationships. The NHS and NIOSH survey systems were not jointly designed and do not provide integrated environmental and health information. Occupational Data Because the occupational environment can be relatively well defined and occupational exposures and resulting disease are theoretically preventable, the workplace is an obvious target for -50-

studying effects of exposure on health outcome. In some cases, occupational groups provide the only sources of data on human health effects from specific chemicals, and this information is used to estimate risk to the general population from exposure to various amounts of the same substances. Essentially, all the comments about epidemiologic studies also apply to the subset of such studies in the occupational sector. From an occupational health surveillance standpoint, an ideal data system would include: a) a listing of all the toxic exposures (including combinations of exposures) that are to be monitored b) adequate workplace measurement of actual exposures, which includes generic names of chemicals and physical agents and data on exposure, duration, and severity c) lifetime exposure records for any worker exposed d) lifetime medical monitoring of exposed workers, including improved physician/hospital reporting, better testing/diagnosis capability, etc. For-~,idable difficulties have hindered the establishment of surveillance systems for occupational illness and disease and death. Many of these difficulties are identical to those encountered in the establishment of surveillance for other environmentally related health problems. Among the problems have been 1) the non-specificity of most occupational illness. Lung cancer caused by asbestos, for example, is not different in any clinical or pathological aspect from that caused by cigarette smoking 2) the long time (latency) that often must elapse between the beginning of occupational exposure and appearance of illness. Few occupational cancers appear within 10 years first exposure, and many do not become evident until after 30 or more years 3) the mixed exposures of most workers to many toxic agents rather than to a single hazard 4) the confounding and occasionally synergistic effects of occupational and other exposures--for example, of asbestos with cigarette smoking. -61-

For these and other reasons, concepts of disease reporting that have been of considerable utility in the control of communicable diseases have found only limited application in occupational disease surveillance. With the institution of a National Death Index in 1981, use of death certificates for epidemiologic studies will become easier, but the certificates still must have relevant information coded dependably. Although states provide uniform reports of causes of death, less than half of the states include occupation and industry information. In an effort to improve the quality of data available on occupation and industry, the National Center for Health Statistics (NCHS), in cooperation with NIOSH and the Census Bureau completed a study in 1979 on the feasibility of coding occupation and industry from death certificates along the lines of the Census Bureau procedures.39 Uniform coding is necessary to insure comparability of the different data sources. NCHS plans to work with states, NIOSH, and the Census Bureau to develop an instruction manual and provide training to the states so that uniform coding of these items can begin to be carried out nationally. Despite many difficulties, occupational diseases and illnesses appear more amenable to surveillance than do health effects caused by non-occupational exposures to hazards in the general environment. The major methodologic advantages of conducting morbidity and death surveillance in occupationally exposed populations are the exposed populations can, on the basis of personnel, Social Security, and other record systems be identified, delimited, and to some extent followed. Further, the exposed populations can frequently be stratified as to either actual measured exposures or such proxy indicators of exposure as job category or duration of employment; 2) occupational populations are frequently more heavily exposed to toxic agents and encounter new toxic substances sooner than does the general publice For those reasons, there is a higher likelihood that any associated toxic effects will occur, and will occur early, in occupational populations . Occupational Survei [lance Occupational survei [lance is best accomplished through a two-tiered system--hazard surveillance plus health effects surveillance. Although numerous potentially useful data sets exist,40 particularly for health effects surveillance, they have not been well integrated, and there remain many gaps. The remainder of this section will describe some of the existing data and their limitations. Recommendations for the establishment of a more nearly -52-

adequate system for national surveillance of occupational hazards, disease, and death will appear in a later section of the report. Hazard surveillance The only nationally comparable worksite hazard identi fication system has been derived from the first National Occupational Hazard Survey (NOHS). It was conducted by NIOSH in 1972-74 by a team of 20 graduate engineers who made observational studies in a probability sample of 5,000 worksites.41 More than 9,000 different potential hazards were discovered, and more than 85,000 trade-name products were listed by these surveyors. Since that survey, more than 10,500 manufacturers of these trade-name products have been contacted, and the chemical ingredients for about 60,000 of the products have been identified. Until this survey was conducted, there was little information available to identify and to inform workers of the hazards to which they were exposed and to direct research and regulatory efforts toward high-priority industrial sectors. Limitations in the sample size and the present age of the data notwithstanding, the NOHS continues to be the main data resource used by government, labor, and medical care providers on the type, extent, and distribution of potential workplace hazards. Data from NOUS I are now at least six years old. Since 1974, there have been substantial changes in the number and types of chemical and physical agents in the workplace, the mix of industries, and the use of control technology. Because of these changes, NIOSH is undertaking a second NOBS that will begin to provide new hazard monitoring data by 1983. A series of reports on special topics and other publications is anticipated from NOHS II over a three- to four-year period when the field phase of the survey is concluded. These publications will examine in detail either a class of chemicals, an industry, or sets of industries and occupations. In addition to NORS II, other projects will contribute to the overall goal of monitoring chemical and physical agents in the workplace. For example: 1) A computerized data file is being built from information on workplace exposures supplied by employees or their unions. Two major unions are cooperating with NIOSH in this project by obtaining from their locals the material safety data sheets (MSDS) and other lis ts of agents used in plants. These fi les can be used to identify potential exposures of union members according to geographic location, plant , and Type of industry. 2) NIOSH is testing a Cooperative Survey Program that can be used by owners and managers of small industries and businesses to carry out a self-inspection that will identify potential workplace hazards. If employers wish, they can send environmental samples to NIOSH for analysis. -63-

3) Trade-name Product Ingredient Clarification File is being developed for use by workers, unions, employers, medical care providers, and occupational health researchers to identify the chemical constituents of trade-name products and possible toxicological risks. Health effects surveillance Monitoring of the impact of occupation on health relies mainly on information systems that were designed for other purposes. At NIOSH a major effort is now under way to work cooperatively with states to build monitoring systems that use mortality and morbidity data compiled by state agencies. It is expected that at least two types of studies will evolve from these data sources. First, proportional mortality and morbidity studies could be done at the state level as advocated by Milham,42 and, second, clusters of unusual health effects could be investigated.43~44 Another occupational health monitoring system makes use of Social Security Administration Disability File data from 1962-72. From these, race-specific, sex-specific, age-adjusted proportional morbidity ratios have been generated for use in identifying possible high-risk occupational groups. A second study will investigate the association of industry groups and cause-specific awards for disability. A third study will deal primarily with the mining industry. In all three studies, the aim is to develop a surveillance method that employs existing data files to pinpoint occupations and industries that appear to warrant further epidemiologic investigation. NIOSH staff also are using data from NCHS' National Health Interview Survey (~HIS) to search for associations between occupation and illness patterns deserving closer study. In the 1980 NHIS, revised questions on occupation and industry provide information that can be used to assess variation in medical care utilization among groups of retired workers for whom lifetime or usual occupation was reported. Information of this type could be applied directly to the assessment of medical care costs associated with environmental (work-related) hazards, as required for the ongoing study. Effects of Acute Environmental Exposures Such episodes as nuclear reactor accidents, chemical spills, and leaks of toxic wastes may cause the acute exposure of populations to environmental contaminants that affect health or result in chronic exposure that raises public concern acutely. Examples of recent environmental emergencies include the nuclear reactor accident at Three Mile Island; a high-dose spill of fluoride into the drinking water system in Michigan;45 a spill of phenol into drinking water in Wisconsin;46 a Mississippi pesticide plant explosion in 1975 that dispersed parathion vapor into the air; and a toxic waste dump fire in Elizabeth, New Jersey.47 The Love Canal chemical dumpsite -54-

in Niagara Falls, New York, and a DDT spill in Triana, Alabama,48 are recent examples of environmental exposures that have led to long-term chronic exposure. The numbers of persons exposed each year to environmental toxins in episodes such as these is not known, but may be large. Although difficult to obtain, statistics collected on exposures and disease outcomes in such persons would provide valuable information for the ongoing study and be particularly responsive to the congressional charge. To evaluate the frequency and the costs of disease and illness associated with acute environmental episodes, two sorts of data must be systematically collected 1) Information must be gathered on the occurrence of acute environmental episodes--on their number, length of exposure, location, and the chemicals or other agents involved. It appears that many of the data necessary for this tabulation already are being collected in an uncoordinated fashion that precludes computation of resulting health effects.49 The U.S. Department of Transportation, for example, collects data on reported spills of toxic chemicals in transit. The Environmental Protection Agency collects data on hazardous dumpsites. The Department of Energy and the Nuclear Regulatory Commission have data on the location and nature of radiation leaks. These data need to be coordinated and analyzed so that estimates can be made of the number of acute environmental episodes that occur each year and of the number of persons exposed or injured as the result of exposure. 2) To evaluate the health effects of acute environmental exposure, the federal government could establish a multidisciplinary acute response team. (The Centers for Disease Control (CDC), with additional resources, might constitute an appropriate administrative locus for such an organization.) It would be helpful if the responsible agency also coordinated health data collection among the many state and federal agencies with interest in the management of toxic environmental emergencies. The function of the response team would be to collect environmental, epidemiologic, and clinical data on acute environmental episodes to enable reliable estimates of the number of persons affected, and the establishment of further determination of dose/response and cause and effect relationships between exposures and subsequent health problems. The acute response team might consist of a core group representing epidemiology, environmental engineering, statistics, and medicine. In addition, there should be available to the team a network of experts in such studies as nuclear physics, chemistry, geology, and toxicology, as well as a network of expert laboratories. -65-

The primary responsibility of this team would be to go rapidly to the sites of acute environmental events in order to conduct studies that would document and measure a) the extent and severity of the environmental insult b) the number of people actually or potentially at risk c) the immediate health effects resulting from the exposure. Tf adverse health effects were found, or if potential future effects were strongly suspected, further study might be required. For example a) cross-sectional evaluation of the entire population at risk or of a statistical sample of the population with additional intensive monitoring of exposures b) establishment of a registry for the prospective longitudinal follow-up of the exposed group. A population-at-risk registry would require the collection of identifying data, information on exposure dose (by environmental, biological, or proxy monitoring), information on baseline health status, and possibly information on potentially confounding life-style factors. Future follow-up of such registries, including follow-up through the National Death Index, would enable future estimation of the incidence of delayed disease resulting from acute environmental exposures. For both the acute and long-term phases of these evaluations, appropriate comparisons must be made to unexposed control groups.50 The results of too many acute environmental studies have been vitiated by failure to employ adequate epidemiologic methodology of this kind. The systematic collection of data on the acute and chronic health consequences of population exposures to acute environmental episodes will constitute an important component of the ongoing data collection activity mandated under P.L. 95-623. These evaluations would provide data, available from no other source, that are necessary to elucidating the high-dose range of the dose/response curves predicting health effects from environmental exposure. Susceptible Populations Many, if not most, factors that contribute interaction between the which the individual is that will determine the well understood illnesses have environmental to their occurrence.] However, it is the environmental components and the degree to susceptible or resistant to these factors severity and outcome of the health effect. -56-

Individuals may be more susceptible or resistant to particular environmental factors because of heredity; their particular developmental stage of life; sex; or acquired conditions, such as nutritional status, pre-existing disease conditions, and, to some extent, personal habits and behavior. Identifying those individuals susceptible to environmental factors can be important in developing occupational and environmental health policies and practices and in interpreting environment-related health statistics. However, the mechanisms responsible for differential susceptibility to toxic environmental agents are known for only a few predisposing conditions and substances. Moreover, attempts to estimate the costs of environment-related health effects for a population usually are based on data derived from studies that assume homogeneous populations. Accordingly, consideration of the effect of various characteristics being evaluated on particularly susceptible groups should be bui It into the ongoing study. Acauired SusceDtibi litY Hei ghtened suscept i hi 1 i ty to envi ronmental a gents may be acquired. 51, 52 For example, suppression of the body's immune system results in an increase in susceptibility to some kinds of microbial agents. Certain chronic diseases, such as bronchitis and asthma, may make some people increasingly liable to the irritating effects of photochemical smog or chemical exhausts in the atmosphere. There are instances in which people develop hypersensitivity to substances to which they are exposed. The phenomenon of acquired susceptibility needs further study to determine its incidence. Inherited Individual Susceptibility - Many of the best known and understood examples of heightened susceptibility are inherited, or probably inherited, and involve changes in single genes. However, other cases of inherited susceptibility are controlled by two or more genes that interact with one or more environmental factors. A mutant gene may have little or no significant effect on well-being until the person is in an environment that stresses a biochemical system influenced by the gene. If this stress is great enough, the cellular integrity of target tissues may be disrupted or damaged badly enough to cause an illness that may be acute or chronic. Some inherited conditions already have been identified and -67-

characterized in which the response to environmental factors and subsequent severity of disease vary according to the type of mutant.^ It is conceivable and even likely that every individual in the population is especially susceptible to some environmental factors. The overwhelming majority of genetic factors or the environmental factors that may contribute to differences in susceptibility are not known. Appendix F lists some inherited conditions that may directly or indirectly influence an individual's risk of developing an illness related to environmental factors. Age and Sex During pregnancy the mother and fetus are particularly susceptible to a variety of environmental factors.53-56 Experiments with rodents indicate that the fetus is more sensitive than the adult to certain carcinogens by several orders of magnitude. Further, the newborn rodent is also more susceptible than the adult, and possibly even than the fetus, to some carcinogens, because it is no longer protected by the maternal body. Various stages in the human life span are associated with increased susceptibility. The developing embryo in the first eight weeks has been reported to have a high susceptibility to certain environmental influences, especially drugs.-53 This is a time when a great majority of birth defects are induced, and the conditions of exposure to certain types of agents are of particular concern in relation to such structural defects. Other groups subject to *For example, a large number of mutations of the gene controlling production of the enzyme glucose-5-phosphate dehydrogenase (G6PD) have been described.! Two of these, an African and a Mediterranean variant, are associated with limited breakdown of red blood cells when males affected by this deficiency take certain drugs, e.g. antimalarial drugs, such as primaquine, or ingest certain chemicals, e.g. naphthalene (the major ingredient in moth balls). However, the severity of red blood cell damage is much greater in those who have the Mediterranean variant than in those who have the African variant. Furthermore, ingestion of the fava bean will damage red blood cells in males who carry the Mediterranean variant of C~5PD deficiency, but will generally have no detectable clinical effect on males who have the African variant. This mutation is quite common, with, for example, in the United States approximately 12 percent of black males of African descent having G6PD deficiency. The proportion of whites of Mediterranean ancestry who have the more severe form varies with different ethnic groups. -68-

increased susceptibi lity because of their age include the young, during the perinatal and breast feeding periods, and the elderly. The issue of individual susceptibility will become more important as information increases. For accurate risk estimates and quantification of costs, groups with special susceptibility need to be considered. Observations often are based on effects seen at relatively high doses in a small population without adequate representation of possibly susceptible groups. Low doses will have their greatest effects on the susceptible groups. Therefore, effects extrapolated from high doses may underestimate or miss effects on susceptible populations. Questions of individual susceptibility raise many additional issues, such as the cost of surveillance of high risk groups, costs of worker's compensation, confidentiality, and the setting of priorities for funding. This committee recognizes that these issues exist and notes that they deserve study and consideration. Effects of Environmental Agents on Human Reproduction The planning committee singled out the effects of environmental agents on human reproduction for special attention. There is growing concern about such effects and a broader range of effects is being noted than in the past, but the kinds of data that are available do not readily lend themselves to studying reproductive effects. Environmental agents may affect reproduction in males and females in many ways. They could affect fertility, diminish libido or sexual function, promote impotence, or otherwise impair reproductive ability by adversely affecting the reproductive cycle or contri but i ng to de fec t ive or i nsu fficient spermatogenesis or oogenesi s . Environmental agents may increase pregnancy . wastage by inducing implantation defects, spontaneous abortions, or stillbirths; they could act as mutagens or teratogens, producing structural, functional, or metabolic defects; they can act as transplacental carcinogens, influence behavioral development, or increase susceptibility to disease. Examples of environmental exposures leading to reproductive effects include abnormalities in sperm and cases of sterility after exposure to dibromochloropropane (~BCP),57 decreased libido after occupational exposure to synthetic estrogens,58 and transplacental carcinogenesis due to in utero exposure to diethylstilbesterol (DES).59 Despite a large accumulation of scientific literature concerning the reproductive effects of environmental agents such as are found in the computerized files of the Environmental Mutagen Information —69—

Center (EMIC)* and the Environmental Teratogen Information Center (ETIC),* the effects of environmental factors on reproduction are not well understood or quantified. In order for the ongoing study to be able to calculate the costs of reproductive effects of environmental agents, a great deal more information will be needed.60 High risk populations, including occupational cohorts, should be identified and monitored for point mutations, chromosome aberrations, abortions, stillbirths, birth weight, and congenital malformations. It would be useful to have sentinel indices and in vitro tests for assessing different levels of reproductive failure and for predicting the pathogenetic potential of environmental agents. *EMIC and ETIC systems are located in Oak Ridge, Tennessee, and they are operated by Union Carbide Corporation under a contract from the U.S. Department of Agriculture. -70-

PEFEPENCES Omenn, G.S. and Motulsky, A.G. Eco-genetics genetic variation in susceptibility to environmental agents. In Genetic Issues in Public Health and Medicine, pp. 83-111. Edited by B.H. Cohen, - A.M. Lilienfeld, and P.C. Huang. Springfield, It CC Thomas, 1978. Land, C.E. Estimating cancer risks from low doses of ionizing radiation. Science 209: 1197-1203, 1980. 3. National Research Council/Institute of Medicine. Food Safety Policy Scientific and Societal Considerations. Washington, D.C.: National Academy of Sciences, March 1979. 4. Hollstein, M., McCann, J., Angelosanto, F.A., and Nichols, W.W. Short-term tests for carcinogens and mutagens. Mutation Research 65~133-226, 1979. U.S. Environmental Protection Agency. Environmental Criteria and Assessment Office. Air Quali ty Cri teri a for Part i curate Matter and Sulfur Oxides. Volume TV Health Effects, April 1980. Draft. Environmental Protection Agency. Availability of TSCA Revised Inventory. Federal Register, pp. 50544-50546, July 29, 1980. 7. Food Safety Counci 1. Proposed System for Food Safety Assessment. Final Report of the Food Safety Council, Washington, D.C., June 1980. 8. Rinkus, S.J. and Legator, M.S. The need for both in vitro and in viva systems in mutagenicity screening. In Chemical Mutagens Principles and Methods for Their Detection. Volume 6. pp. 365-473. Edited by F. de Serres and A. Hollaender. New York Plenum Publishing Corporation, 1979. 9. Ames, B.N. Identifying environmental chemicals causing mutations and cancer. Science 204:587-593, 1979. 10. de Serres, F.J. and Ashby, J., editors. Evaluation of Short-term Tests for Carcinogens Report of the International Co1laborative Program. New York E1 sevi er Press. to be publ i shed 1981. 11. U. S. Department of Health, Education, and Welfare . National Toxicology Program. National Toxicology Program Annual Plan for Fi seal Year 1980. Washin~ton, {).C. November 1979. -71-

12. International Agency for Pesearch on Cancer. Chemicals and Industrial Processes Associated with Cancer in Humans. IARC - Monographs ~ Volumes 1-20, Supplement 1. 1979. Lyon: IAPC, September 13. U.S. Department of Health, Education, and Welfare . National Cancer Institute. Guidelines for Carcinogen Bioassay in Small Rodents. Technical Report Series No. 1. DREW Publication No. (NIH)76-01. Bethesda, MD: National Cancer Institute, 1976. 14. Interagency Regulatory Liaison Group. Scientific Bases for Identifying Potential Carcinogens and for Estimating Their Risks. Washington, D.C.: IPEG, February 6, 1979. 15. Staffa, J.A. and Mehlman, M.A., editors. Innovations in cancer risk assessment (EDo1 study). Journal of Environmental Pathology and Toxicology 3:1-246, 1980. 16. Hates, R.~. Risk Assessment of Environmental Hazards. New York John Wiley and Sons, 1978. 17. Staffa, J.A., editor. Risk/Benefit Decisions and the Public Health. Fockville, MD Food and Drug Administration, 1980. 18. Pall, D.P. Validity of extrapolation of results of animal studies to man. In Public Control of Environmental Health Hazards. Edited by F.C. Hammond and I.J. Selikoff. Annals of the New York Academy of Sciences 399:85-91, 1979. 19. Crouch, F. and Wilson P. Interspecies comparison of carcinogenic potency. Journal of Toxicology and Environmental Health 5~1095-1118, 1979. 20. MacMahon, B. and Pugh, T.F. Epidemiology: Principles and Methods. Boston Little, Brown and Company, 1970. 21 . Li lienfield , A.M. Foundations of Epidemiology. New York Oxford University Press, 1976. 22 . Miller, R.W. The discovery of human teratogens, carcinogens and mutagens lessons for the future. In Chemical Mutagens Volume 5., pp. 101-126. Edited by A. Hollaender and F.J. de Serres. New York Plenum Publishing Corporation, 19 7 8 . 29. Miller, R.W. Area-wide chemical contamination lessons from case histories. Journal of the American Medical Association. To be publi shed. 24. U. S. Environmental Protection A~encv. ~ , Health Consequences of Sulfur Oxides A Report from CHESS, 1970-1971 e EPA-550/1-74-004 e May 1974 e -72- .\

Speizer, Em., Bishop, Y., and Ferris, R.G. An epidemiologic approach to the study of the health effects of air pollution. In Symposium on Statistics and Environment. pp. 56-58. Held 1976e Washington, D.C.: American Statistical Association, 1977. 26. Spengler, J.D. and Dockery, D.W. Long-term measurements of respirable sulfates and particles inside and outside homes. Paper delivered at session on Environmental Monitoring and Assessment. Electro Pro fe s s i anal Programs . New York . April 24-25, 1979. 27. Newhouse, M.J,. and Thompson, H. Epidemiology of mesothelial tumors in the London area . Annal s of the New York Academy of Sciences 132: 579-588, 1965. 28. Beebe, G.W. Reflections on the work of the Atomic Bomb Casualty Commission in Japan. Epidemiologic Reviews 1~184-209, 1979. Fox, A.J. Prospects for measuring changes in differential mortality. WHO organization paper given at Meeting on Socioeconomic Determinants and Consequences of Mortality, Mexico City, June 19-25, 1979. 30. Fox, A.J. The role of OPCS in occupational epidemiology. Annals of Occupational Hygiene 21 393-403, 1979. 31. Packney, J.D. and Linn, W.S. Koch's postulates updated: a potentially useful application to laboratory research and policy analysis in environmental toxicology. American Review of Respiratory Disease 119:849-852, 1979. 32. U.S. Department of Health and Human Services. Environmental Health A Study of the Issues in Locating, Assessing, and Treating Individuals Exposed to Hazardous Substances. DUBS Publication No. (PHS)81-1975. Washington, D.C.- U.S. Government Printing Office, 1981. 33. National Heart, Lung, and Blood Institute. Epidemiology and biometry program. NPLBI Fact Book--Fiscal Year 80. Department of Health and Human Services. Washington, D.C., 1980. 34. Kannel, W.B., Castelli, W.P., and Gordon, T. Cholesterol in the prediction of atherosclerotic disease. Annals of Internal Medicine 90 85-91, 1979. 35. Doll, P., Payne, P., and Waterhouse J Five Continents. Volume I. UICC-International Union Against , = . Cancer. New York: Springer-Verlag, 1966. —7 3—

36. Case, R.A.M., McDonald, D.B., and Pearson, J.T. Tumors of the urinary bladder in workmen engaged in the manufacture and use of certain dyestuffs intermediates in the British chemical industry, Part I. The role of aniline, benz;dene, alpha-naphthylamine and beta-naphthylamine. British Journal of Industrial Medicine 11:7S-104, 1954. - 37. National Research Counci1. Environmental Studies Board. Environmental Monitoring. Washington, D.C. ^ National Academy of Sciences, 1977. 38. National Center for Health Statistics. Data Systems of the National Center for Health Statistics. Publication No. .. . . . . (PHS)80-1247. Hyattsville, MD National Center for Health Statistics, January 1980. 39. Rosenberg, H., Burham, D., Spirtas, ~ ., and Valdi sera, V. Occupation and industry information from the death certificate assessment of the completeness of reporting. In the Proceedings of the American Statistical Association, August 1979. 40. Discher, D.P., Raymond, M.A., and Witwer, C.R. Evaluation of existing data sources of occupational injury and illness statistics. SRI Project 4434, Stanford Research Institute, June 1976. 41. U.S. Department of Health, Education, and Welfare. National Occupational Hazard Survey. Volumes 1-3. DREW Publication No. (NIOSH)74-127. May 1974. 42. Milham, D., Jr. Occupational Mortality in Washington State, 1950-71. Volume III. DREW Publication No. (NIOSH)76-175-A-C. Washington, D.C.. 1976. 43. Rutstein, D.C., Berenberg, W., Chalmers, T.C., et al. Measuring the quail ty of medical care a clinical method. New England Journal of Medicine 94 582-588, 1976. 44. Frazier, T.M. and Wegman, D.H. Exploring the use of death certificates as a component of an occupational health surveillance system. American Journal of Public Health 69:718-720, 1979. 45. Center for Disease Control Internal Memorandum EPI-78-24-2. Fluoride Poisoning in Michigan. Atlanta: Center for Disease Control, December 12, 1978. 46. Baker, E.L., Landrigan, P.J., Bertozzi, P.E., et al. Phenol poisoning due to contaminated drinking water. Archives of Environmental Health 33 89-94. 1978. -74-

47. Halperin, W.H., Landrigan, P.J., Altman, R., et al. Chemical fire at a toxic waste disposal plant: an epidemiologic study of smoke and fumes. New Jersey State Medical Journal, in press. 48. Center for Disease Control Internal Memorandum EPI-79-44-2. Possible Community-wide DDT Contamination, Triana, Alabama. Atlanta: Center for Disease Control, October 15, 1979. 49. U.S. Department of Commerce. Office of Federal Statistical Policy and Standards. A Framework for Planning U.S. Federal Statistics for the 198Os. Chapters 8, 17, and 26. Washington, D.C. U.S. Government Printing Office, July 1978. 50. Proceedings of an International Workshop on Plans for Clinical and Epidemiologic Follow-up After Area-wide Chemical . . Contamination. Held March 17-19, 1980, Washington, D.C. Sponsored by Assembly of Life Sciences, National Academy of Sciences. To be published. 51. From the NIR occupational asthma from a low-molecular-weight organic chemical. Journal of the American Medical Association 244 1667-1668' 1980. 52. Check, W.~. Work-related asthma comes out of hiding. Journal of the American Medical Association 244:1653-1659, 1980. 53. Leck, I. Teratogenic risks of disease and therapy. to Epidemiology and Biostatistics 1:23-43, 1979. 54. The John E. Fogarty International Center for Advanced Study in the Heal th Sciences . Prevention of Embryonic, Fetal, and Perinatal Disease. DREW Publication No. (NIH)76-853. Washington, De C.: U.S. Government Printing Office, 1976. 55. Weiss, B. and Spyker, J.M. Behavioral implications of prenatal and early postnatal exposure to chemical pollutants. Pediatrics 53:851-859, 1974. 56. Society for Occupational and Environmental Health. Proceedings: Conference on Women and the Workplace. Washington, D.C.: Society for Occupational and Environmental Health, April 1977. Wharton, D., Milley, T.H., Krauss, R.M., and Stubbs, H.A. Testicular function in DBCP exposed pesticide workers. Journal of Occupational Medicine 21 161-166, 1979. 58. Stellman, J.M. The effects of toxic agents on reproduction. Occupational Health and Safety 48:36-43, 1979. -7~

59. Herbst' A.L., Ulfelder, H., and Poskanzer, D.C. Adenocarcinoma of the vagina. New England Journal of Medicine 284:878-881, 1971. 6Q. Buffler, P.A. Some problems involved in recogulzing teratogens used in industry. ~ y_~nd Blostatistics 1:118-137, 1979 ,, -76-

Next: Chapter 4: Costs of Health Effects »
Cost of Environmental-Related Health Effects: A Plan for Continuing Study Get This Book
×
Buy Paperback | $50.00
MyNAP members save 10% online.
Login or Register to save!
  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!