IV. METHODS

Epidemiology is basic to modern public health. It provides, for example, the rational basis for health planning, the justification for allocating funding, and the basis for deciding whether or not to introduce or change preventive health policies. Finally, it plays a fundamental role in making decisions concerning optimal treatment regimens through its involvement in the clinical evaluation process.

Epidemiology is distinct from medical science in that epidemiology's focus is on population health, opposed to medicine's focus on the individual patient. While medicine seeks to heal the individual who, by virtue of being susceptible, becomes ill, epidemiology seeks to identify the underlying cause that results in illness among those who are susceptible. With an underlying cause identified, it becomes possible to intervene at the source of the chain of events that leads to illness among people who are susceptible. Removal of the cause can directly result in preventing those who are susceptible from being exposed to it in the first place and thereby from becoming ill.

Epidemiology focuses on large numbers of people comprising populations or communities. It is a quantitative (as opposed to a qualitative) science whose methods are heavily dependent on the application of biostatistical principles and on advances in biostatistical methods. As with other quantitative sciences, epidemiology requires the counting, classification, and analysis of sizable amounts of data. In order to derive meaning from large amounts of data, statistical techniques are used to produce various kinds of summaries. These techniques are known as biostatistics in the health and/or biological sciences.

Through the early 1940s, prior to the advent of antibiotics at the time of World War II, epidemiologists were occupied almost exclusively with controlling infectious diseases. Success resulted in better control of infectious diseases; improved living standards, especially in developed countries; and increased life expectancy of the population. Consequently, epidemiology expanded from its preoccupation with infectious diseases to include noninfectious diseases.

The notion that noninfectious and, by extension, chronic diseases can be prevented by eliminating their causes, analogous to the prevention of infectious diseases, is a relatively new concept. Hence, the modern role of epidemiology, from the public health perspective, is to identify appropriate interventions for consideration by policymakers for controlling disease at the source and thereby promoting health in the community.

The linking of epidemiology and biostatistics has become a hallmark of modern epidemiology in both its research and its practice areas of activity. Research in epidemiology tends to embrace activities of an experimental nature, while the practice domain tends to focus on disease surveillance and monitoring activities. Regardless of the domain, biostatistics provides the analytic tools used in epidemiology.

Scientific discovery in the laboratory should ultimately have practical application at the bedside. Results of epidemiologic investigations made on a population or on clearly defined subgroups of the population ought to benefit individuals. Because the results of population-based research are couched in terms of probabilities, the application of epidemiologic studies to the individual is not direct. Nevertheless, the identification of risk factor information in the absence of a biologically identified cause of a disease has been instrumental for prevention programs. Furthermore, physicians can apply probabilities in deciding therapeutic options.

The Scope of Epidemiologic Activity

Epidemiologic studies are necessary to provide both valid and reliable data not only concerning the distribution of diseases in populations, but also on the impact of social, economic, environmental, and other factors on the health of populations. In addition, epidemiologic data are often fundamental in making future projections of disease burden, crucial for planning purposes.

Concerning professional ethics, in the physician–patient "medical" relationship, the physician assumes a patient advocacy role; epidemiologists, on the other hand, assume a population/community advocacy role. Ethical guidelines that have been developed for medicine therefore have little relevance to epidemiology. Obligations assumed under these two different models must be explicit for trust to exist between professionals and the public.

Since the 1960s, epidemiology has undergone dramatic growth, paralleling to some extent the growth and development of computers. In North America, for example, the sex distribution and training of epidemiologists has changed over this period. Previously, epidemiologists were predominantly male, but today about half, especially those engaged in research, are women. Also, about half of today's epidemiologists were never trained as physicians.

The absolute numbers of epidemiologists have grown exponentially and the development of advanced computer technology has enabled epidemiologists to work with and share increasingly larger databases and to apply sophisticated multivariate statistical adjustment techniques via the use of computer software. But while technology has led to important advances in epidemiology, the complex issues of ensuring both integrity in science and ethical conduct among scientists have yet to be adequately addressed. There is increasing recognition of the need for guidelines to ensure professional accountability to the public in whose service epidemiologists work.

Classical epidemiology—as distinct from clinical evaluation—is primarily an observational science; it studies the events of daily life among the members of the various subgroups that comprise a community. Unlike controlled experiments, epidemiologic research measures events associated with populations whose lifestyles, work habits, and other characteristics have evolved outside the epidemiologist's control. Because uncontrollable and unknown risk factors can impact study outcomes radically, they must be accounted for if demonstrated contrasts, comparisons, and differences are attributed to these. Epidemiologic methods include various approaches for ensuring appropriate analysis of observational events. Professional epidemiologists are cognizant of the strengths, weaknesses, and limitations of the various methodologic options in light of the complexities associated with the conduct of uncontrolled experiments.

The closest epidemiology comes to the conduct of a controlled experiment is in the randomized controlled trial (RCT). However, RCTs can be justified only on the basis of substantial preexisting information concerning the intervention of interest (e.g., a particular therapy). Preexisting information usually is derived from the conduct of studies utilizing designs that are nonexperimental in nature (i.e., from the realm of natural experiments). Only where justification exists can human beings be subjected to random allocation in a clinical trial. Natural experiments in observational research include descriptive, ecological, retrospective case-control, and prospective cohort designs.

Diseases associated with aging, including cancer, diabetes, and cardiovascular diseases, have required greater attention. Because epidemiology provides the methodology for rational approaches to interventions, epidemiology is fundamental to disease prevention. Interventions based on epidemiological studies have taken the form of health promotion programs, such as campaigns for smoking cessation, no drinking and driving, and condom use in sexual inter-course. The onset in 1981 of the acquired immunodeficiency syndrome (AIDS) pandemic, however, reminded epidemiologists that infectious diseases are not necessarily a thing of the past.

With escalating healthcare costs in Canada, the United States, and elsewhere, epidemiology is playing a major role through providing the evaluative methodology for assessing cost-effective interventions for rational healthcare planning. Epidemiologists establish health goals by assessing health status indicators for a population; they identify target levels for reduced morbidity, disability, and mortality. These activities have implications for resource allocation which bear directly on the ethical principle of distributive justice. Indeed, numerous jurisdictions are attempting to identify those illnesses for which free health coverage should be provided by the "state" based on prevailing population values. Epidemiology assists in these determinations through expertise in survey methodology, health-status indicators, and disease classification.

From the foregoing, it is clear that epidemiology plays a major role in health-policy decisions, which involve, among others, substantial financial resources. "Health" is big business. Concerns arose during the 1980s about the possible influence of individuals and/or groups whose vested interests could bias outcome(s), motivated by financial profit and/or professional prestige. Conflicting-interest issues have been of concern not only in the interpretation of epidemiologic studies in favor of any one interest group's position but even in limiting or blocking the potential to conduct the best possible epidemiologic study for addressing a health concern.

The legal aspects—in terms of civil, administrative, and criminal law—are profound. With utilitarian goals in mind(i.e., doing the greatest good for the largest number of people), the courts usually have invoked the collective good over individual freedoms (e.g., in legislation concerning vaccination, quarantine, seat belts, and smoke-free public indoor environments in both Canada and the United States).

In general, governments prefer that professions regulate themselves. Professional organizations are expected to do what is necessary to minimize scientific misconduct and ensure professional etiquette among employers, sponsors, colleagues, and clients.

A Historic and Ongoing Concern: Privacy

Any epidemiologic investigation conducted under the auspices of an institution (e.g., a university, hospital, or government office) is likely to be subjected to ethical review by a committee. The committee usually comprises members of various disciplines as well as a lay representative.

Not only can ethics review committees examine the nature of the question to be addressed by the investigation, but they also may determine the appropriateness of the methods being proposed. Generally, however, the main focus tends to be on the possible harms versus benefits to those who will participate in the study; that is, with issues of privacy, informed consent, and confidentiality, and most important, that none of the procedures expected of the subjects/participants will cause them harm.

Scientific peer review concentrates on the aptness of the proposed scientific research methods, including the scientific relevance of the proposed research question, assessment of potential bias and confounding, adequacy of the proposed size of the study and associated statistical power, and recognized limitations impacting on the interpretation of the study. These two distinct but related areas of concern are seldom brought to the attention of a single expert other than the principal investigator, and perhaps also his or her research team. Without the support of both groups, the proposal usually cannot proceed into action.

Because the data epidemiologists rely on can be personally sensitive, governments have enacted privacy legislation to protect its citizens. Only with special permission from the custodians of these data bases can epidemiologists gain access—usually controlled—to the data banks essential to the conduct of health research. Some agencies also impose an oath of secrecy on the researcher.

One protection that researchers are expected to exercise (in their publication of results from access to health records in the public domain) is the anonymity of all persons studied. In addition, the identification of small areas or groups of people must be avoided also to ensure anonymity and thereby the protection of individual privacy. Individual or group stigmatization is to be avoided. Any infringement of the public trust could have repercussions, including legal penalties to the researcher involved. Furthermore, the epidemiologic research enterprise could be placed in jeopardy by engendering a loss of trust in research by the very communities whose support (both financial and possibly also participatory) is needed for investigation purposes.

Professional training, in conjunction with well-publicized guidelines, is likely to minimize any risk of infringement. In addition, the epidemiologist has an obligation not only to respect the right to privacy of personal data, but to ensure that co-workers are equally vigilant. "Whistleblowing" also must be encouraged and those doing so must be protected from any form of reprisal. Most professional ethics guide-lines/codes require that attention be drawn to the person who elects to perform contrary to normative standards of professional practice.

In 1991, European Community government officials developed a set of proposals concerning rights to privacy. Unfortunately, if enacted, these proposals could serve to make it virtually impossible to conduct epidemiologic research that depends on access to these data banks. The proposals ensure that personal information provided for one purpose cannot be used for another purpose without prior consent. Similar legislative proposals were mounted in the United States in the mid-1970s, but were defeated. Hence, epidemiologists and biostatisticians worldwide have a duty to remain vigilant of legislative proposals that might, directly or indirectly, adversely impact research for the public's health. They must be organized enough to provide input to such legislative proposals. Ultimately, it is the public-health interest that must prevail.

Current Issues

ETHICS GUIDELINES. The first stated need for guidelines on the ethical conduct of epidemiologists was printed in 1985. Despite considerable debate within the profession in North America, through 1987, little movement was made. It was at the International Epidemiological Association's (IEA) 1987 XIth Scientific Meeting in Helsinki, Finland, that the proposal to develop guidelines was adopted. By 1990, further discussion had advanced the thinking on this subject and a first draft of IEA guidelines was published.

A milestone conference on the subject of ethics in epidemiology had stimulated the discussion in 1989. The conference had been organized by the United States' Industrial Epidemiology Forum. The organizers had compiled a set of ethics guidelines and a commentary; these subsequently were published in the conference proceedings in 1991. Since then, the Council for International Organizations of Medical Sciences (CIOMS) has published International Guidelines for Ethical Review of Epidemiological Studies together with a compendium conference proceedings which contributed to the development of these guidelines. In addition, CIOMS published International Ethical Guidelines for Biomedical Research Involving Human Subjects. (CIOMS, 1991, 1993).

In November 1991, the American College of Epidemiology was accorded the leadership role among the North American epidemiology bodies to further ethics initiatives in this region of the world. Other groups of epidemiologists with specialty interests are contributing to this process (e.g., environmental epidemiologists).

The Industrial Epidemiology Forum's Guidelines, modeled on those developed some years earlier by the International Statistical Institute, are organized as follows:

I. Obligations to the subjects of research

to protect their welfare, ensuring no physical or mental harm through their participation;

to obtain their informed consent, ensuring the fullest possible understanding of any risks and benefits associated with participation;

to protect their privacy, ensuring no stigmatization resulting from information provided through their participation;

to maintain confidential information, ensuring the privacy of the participant.

II. Obligations to society

to avoid conflicting interests, recognizing that vested interests could bias research in ways that fail to serve the goal of seeking truth;

to avoid partiality by openly recognizing one's biases;

to widen the scope of epidemiology by teaching its methods to interested candidates;

to pursue responsibilities with due diligence;

to maintain public confidence in the profession by ensuring that both the strengths as well as the limitations of the profession are disclosed.

III. Obligations to funders and employers

to specify obligations, ensuring that the values and principles to which epidemiologists are expected to abide are clearly understood;

to protect privileged information, respecting the need of employers and providers of information to have reasonable time to assess the implications of research utilizing their data to their interests prior to disseminating the results from such a study.

IV. Obligations to colleagues

to report methods and results for wider peer review;

to confront unacceptable behavior and conditions, ensuring ethical conduct in support of the public interest;

to communicate ethical requirements, thereby ensuring accountability of the profession to the public.

Loreen Herwaldt (1993) has extended the guidelines set forth by the Industrial Epidemiology Forum by identifying principles having special relevance to hospital infection control officers and clinical practice.

While guidelines, commentaries, and case studies are recognized as essential to ethical conduct, they are insufficient. They must be taught, learned, discussed, challenged, and revised in light of case studies, if they are to affect behavior. Finally, mechanisms for dealing with allegations of breaches of conduct need to be established with remedies that serve to mitigate any wrongs.

CONFLICTING INTERESTS. Objectivity is required both on the part of the epidemiologist who is proposing a research project or submitting a manuscript for publication and on the part of the scientific peer review committee members. A conflict of interest arises when a reviewer has a vested interest in the subject under review that can either positively or negatively impact on the review decision. When a reviewer has a conflict of interest—whether at the scientific approval stage, the ethics review stage, or the publication stage—this must be declared and such reviewer's comments should be considered in this light in any final decision.

Reviewers have an obligation never to use, or to discuss with others, the ideas conveyed in a proposalmanuscript without full attribution to the person who proposed them. To do otherwise would misappropriate the intellectual property of another. In addition, if the reviewer is in a position to execute another's proposal, whether funded or not, such work should not proceed without the prior written permission of the person whose idea it was.

SCREENING FOR DISEASE AND HIV ANTIBODY. As a means of secondary prevention, early detection of disease through screening programs is well recognized. The AIDS pandemic, however, has presented new challenges well documented by Ronald Bayer and his colleagues, whose concern has been more with the stigmatization of individuals or groups. Access to test results by, for example, employers, landlords, or insurance companies has been of concern to infected people who fear job or housing loss and noninsurability. In research involving sexual practices, for example, the investigator requires special legal protection not only to render data inaccessible under subpoena but also to disclose such issues as the sexual abuse of children to child welfare authorities. Since valid responses must be obtained from persons volunteering for research if epidemiologic studies are to be useful, the right to privacy by the person being studied has to be secured in order for the person to participate honestly in the study.

In its initial years, testing for the human immunodeficiency virus (HIV) antibody was intended (together with self-exclusion) to secure the safety of the donated blood supply. Shortly thereafter, however, there were mandates for the testing of population subgroups believed to be at high risk of infection. It was postulated that the HIV antibody test could separate those truly positive from those truly negative, after which one could identify or physically separate the positives from the negatives. (The Cuban model, applied since early in the epidemic, has required that all persons found to be HIV-antibody positive be confined to a common residence and thus be barred from associating with persons who are not HIV-antibody positive.) Unfortunately, no test provides 100 percent sensitivity and specificity for HIV antibody or any other test. Furthermore, a "window period" exists between time of exposure and infection with HIV and the actual development of antibody. This window period can range from about three weeks to several months during which time the individual would test negative when in fact he or she could transmit the virus. This example demonstrates how epidemiology can assist in the rational presentation of facts, thus preventing misinterpretation by the media and/or lobby groups not fully informed of the scientific facts and how to interpret them.

NOTIFICATION. When special subgroups are identified for a study, the results of that study should be provided to the participants. Specifically, in occupational cohort studies, it is recommended in the United States that study participants be informed of any exposure to health risks uncovered through the study. The question that remains relates to the welfare of other workers who may be exposed to similar risk factors and who therefore could be at the same level of risk as those workers who were actually studied. If the cohort study that initially identified the risk was well-designed, it might be possible to extrapolate the research findings to other subgroups at risk in similar occupations, as well as to former employees. These latter two potentially at-risk groups are not currently included in the United States' National Institute for Occupational Safety and Health (NIOSH) guidelines.

Technologies continue to grow for determining individual susceptibility to illness that arises from workplace exposure to hazardous substances. If employers were privy to such information, they could exclude a job applicant on the grounds of wishing to protect the individual and at the same time to protect themselves from potential litigation. The tension arises between the obligation for full disclosure by the job applicant/worker on the one hand, and the obligation of the employer to provide a safe workplace. Some employers have argued that to render a workplace safe could be economically impractical. The controversy continues. Women, for example, face restrictions on employment in certain industries for fear by employers of liability—based on the existing body of knowledge about exposure to certain substances during pregnancy—if pregnancy should result in any abnormality at birth.

One mechanism for disseminating information involves community participation at all stages of a study, from hypothesis formulation through proposal development, review, conduct, analysis, write-up, and interpretation. In this way, community values are integrated into the research. Some occupational health studies have succeeded simply by establishing steering committees. These include not only scientists but also labor and management. Government involvement on a steering committee may also be appropriate.

WOMEN AND MINORITIES. The U.S. National Institutes of Health has stated that research has focused disproportionately on white male subjects (Dresser). Results from studies on males are generalized to other population subgroups (i.e., to women and racial minorities) when the results, in fact, may not be generalizable. Such inferences may not only be misleading for the health of women and minorities but also could create harm through the potentially inappropriate application of findings from studies on white males to other groups in the United States. Therefore, it has now been mandated in the United States that women and minorities be included in all research programs whenever possible (NIH/ADAMHA).

It is difficult to quarrel with the concerns and remedies noted above. However, epidemiology is undertaken in populations not only where the problem to be investigated arises but also in populations that are large enough to satisfy statistical considerations. That is, access to exposed populations is what motivates and justifies epidemiologists to design and conduct a study. Statistical power is a function of the prevalence of exposure in a population. If a large enough number of women or minorities is not exposed to a given agent (e.g., chemical or pathogen) of interest, then their inclusion in studies could be unproductive, consequently wasting resources. Clearly, the researcher must be cognizant of the limits to which inferences can be drawn from any study; it is up to those formulating policy, however, to provide the incentives needed to encourage and enable the address of researchable questions of relevance to groups other than white males.

Assessment to Date and Future Directions

Only recently have ethics guidelines been drafted for epidemiologists, whereas statisticians had broached the subject and developed guidelines in the 1980s. Physicians have been concerned with professional standards of practice in North America since the late nineteenth century. Although epidemiologists indeed may be entering the ethics discussion later than their counterparts, the relative recency of the profession must, of course, be considered. In their favor, epidemiologists are making efforts not only to develop ethics guidelines but also to integrate ethics into their teaching programs and into continuing professional education more generally. Ultimately, the expectation is that grass-roots involvement will maximize the likelihood of adherence to guidelines; the greater accountability of the profession to the public in whose interest epidemiology functions will be more assured.

Of growing concern are issues of self-interest and conflicting interests that sometimes take precedence over the public interest. Greater attention is being given to the consequences of research for destructive purposes through possible harm to the ecosystem and the advancement of militarism. Unless the professions are conversant with the principles of ethics, technological advances will continue to outstrip the ability of professions to respond; the professions' role will continue to be one manifesting a reactive as opposed to a proactive position.

colin l. soskolne (1995)

bibliography revised

SEE ALSO: AIDS: Public Health Issues; Coercion; Conflict of Interest; Economic Concepts in Healthcare; Epidemics; Information Disclosure, Ethical Aspects of; Informed Consent; Minorities as Research Subjects; Occupational Safety and Health; Privacy and Confidentiality in Research; Profession and Professional Ethics; Public Policy and Bioethics; Research, Unethical; Technology;Whistleblowing; and other Public Health subentries

BIBLIOGRAPHY

Bankowski, Z.; Bryant, John H.; and Last, John M. 1991. Ethics and Epidemiology: International Guidelines. Proceedings of the XXVth CIOMS Conference, Geneva, Switzerland, 7–9 November, 1990. Geneva: Council for International Organizations of Medical Sciences.

Bayer, Ronald. 1993. "The Ethics of Blinded HIV Surveillance Testing." American Journal of Public Health 83(4): 496–497.

Bayer, Ronald; Dubler, Nancy N.; and Landesman, Sheldon. 1993. "The Dual Epidemics of Tuberculosis and AIDS: Ethical and Policy Issues in Screening and Treatment." American Journal of Public Health 83(5): 649–654.

Bayer, Ronald, and Fairchild-Carrino, Amy. 1993. "AIDS and the Limits of Control: Public Health Orders, Quarantine, and Recalcitrant Behavior." American Journal of Public Health 83(10): 1471–1476.

Bayer, Ronald, and Toomey, Kathleen E. 1992. "HIV Prevention and the Two Faces of Partner Notification." American Journal of Public Health 82(8): 1158–1164.

Beauchamp, Tom L.; Cook, Ralph R.; Fayerweather, William E.; Raabe, Gerhard K.; Thar, William E.; Cowles, Sally R.; and Spivey, Gary H. 1991. "Ethical Guidelines for Epidemiologists." Journal of Clinical Epidemiology 44 (suppl. 1): 151S–169S.

Coughlin, Steven S., and Beauchamp, Tom L., eds. 1998. Ethics and Epidemiology. New York: Oxford University Press.

Council for International Organizations of Medical Sciences (CIOMS). 1991. International Guidelines for Ethical Review of Epidemiological Studies. Geneva: Author.

Council for International Organizations of Medical Sciences (CIOMS). 1993. International Ethical Guidelines for Biomedical Research Involving Human Subjects. Geneva: Author.

Darragh, Martina, and McCarrick, Pat Milmoe. 1998. "Public Health Ethics: Health by the Numbers." Kennedy Institute of Ethics Journal 8(3): 339–358.

Dresser, Rebecca. 1992. "Wanted: Single, White Male for Medical Research." Hastings Center Report 22(1): 24–29.

Elston, Robert C.; Olson, Jane M.; and Palmer, Lyle, eds. 2002. Biostatistical Genetics and Genetic Epidemiology. Hoboken, NJ: John Wiley & Sons.

Fawcett, Eric. 1993. "Working Group on Ethical Considerations in Science and Scholarship." Accountability in Research 3: 69–72.

Fayerweather, William E.; Higginson, John; and Beauchamp, Tom L., eds. 1991. "Industrial Epidemiology Forum's Conference on Ethics in Epidemiology." Journal of Clinical Epidemiology 44 (suppl.). Special issue.

Gordis, Leon, and Gold, Ellen. 1980. "Privacy, Confidentiality, and the Use of Medical Records in Research." Science 207(4427): 153–156.

Gordis, Leon; Gold, Ellen; and Seltser, Raymond. 1977. "Privacy Protection in Epidemiologic and Medical Research: A Challenge and a Responsibility." American Journal of Epidemiology 105(3): 163–168.

Herwaldt, Loreen A. 1993. "National Issues and Future Concerns." In Prevention and Control of Nosocomial Infections, 2nd edition, ed. Richard P. Wenzel. Baltimore: Williams & Wilkins.

Hoffman, Richard E. 1984. "The Use of Epidemiologic Data in the Courts." American Journal of Epidemiology 120(2): 190–202.

Jekel, James F.; Katz, David L.; and Elmore, Joann G. 2001. Epidemiology, Biostatistics, and Preventive Medicine. Philadelphia, PA: W B Saunders.

Jowell, Roger. 1986. "The Codification of Statistical Ethics." Journal of Official Statistics 2(3): 217–253.

Lachmann, Peter J. 1998. "Public Health and Bioethics." Journal of Medicine and Philosophy 23(3): 297–302.

Lappe, Marc. 1986. "Ethics and Public Health." In Maxcy-Rosenau Public Health and Preventive Medicine, 12th edition, pp. 1867–1877, ed. John M. Last. Norwalk, CT: Appleton-Century-Crofts.

Last, John M. 1987. "Ethical Issues in Public Health." In Public Health and Human Ecology, pp. 351–370. East Norwalk, CT: Appleton & Lange.

Last, John M. 1991. "Guidelines on Ethics for Epidemiologists." International Journal of Epidemiology 19(1): 226–229.

Lilienfeld, Abraham, and Lilienfeld, David E. 1982. "Epidemiology and the Public Health Movement: A Historical Perspective." Journal of Public Health Policy 3(2): 140–149.

Mann, Jonathan M. 1997. "Medicine and Public Health, Ethics and Human Rights." Hastings Center Report 27(3): 6–13.

National Institutes of Health. 1991. "NIH/ADAMHA Policy Concerning Inclusion of Women in Study Populations." NIH Guide 20(32): 1–3.

Oleske, Denise M., ed. 2001. Epidemiology and the Delivery of Health Care Services: Methods and Applications. New York: Kluwer Academic Publishers.

Rose, Geoffrey. 1985. "Sick Individuals and Sick Populations." International Journal of Epidemiology 14(1): 32–38.

Rose, Geoffrey. 1989. "High-Risk and Population Strategies of Prevention: Ethical Considerations." Annals of Medicine 21(6): 409–413.

Rosen, George. 1958. A History of Public Health. New York: MD Publications.

Rothman, Kenneth J. 1981. "The Rise and Fall of Epidemiology, 1950–2000 a.d." New England Journal of Medicine 304(10): 600–602.

Russel, Elizabeth, and Westrin, Claes-Goran. 1992. "Ethical Issues in Epidemiological Research." In Medicine and Health: Workshop on Issues on the Harmonization of Protocols for Epidemiological Research in Europe, Florence, June 30 to July 2,1991. ed. Manuel Hallen and K. Vuylsteek. Commission of the European Communities, COMAC Epidemiology EUR 14596 EN.

Schulte, Paul A. 1991. "Ethical Issues in the Communication of Results." Journal of Clinical Epidemiology 44 (suppl. 1): 57S–61S.

Severson, Richard K.; Heuser, Linda; and Davis, Scott. 1988. "Recontacting Study Participants in Epidemiologic Research." American Journal of Epidemiology 127(6): 1318–1320.

Soskolne, Colin L. 1985. "Epidemiological Research, Interest Groups, and the Review Process." Journal of Public Health Policy 6(2): 173–184.

Soskolne, Colin L. 1986. "Scientific and Ethical Conflicts in Cancer Studies Involving Human Subjects." Women and Health 11(3–4): 197–215.

Soskolne, Colin L. 1989. "Epidemiology: Questions of Science, Ethics, Morality, and Law." American Journal of Epidemiology 129(1): 1–18.

Soskolne, Colin L. 1991. "Ethical Decision-Making in Epidemiology: The Case Study Approach." Journal of Clinical Epidemiology 44 (suppl. 1): 125S–130S.

Soskolne, Colin L. 1991–1992. "Rationalizing Professional Conduct: Ethics in Disease Control." Public Health Reviews 19(1–4): 311–321.

Soskolne, Colin L. 1992. "Reader Questions Extensive Funding for Women's Health." New Epidemiology Monitor 13(10): 7.

Soskolne, Colin L. 1993a. "Ethics and Policy Lost in Headline." New Epidemiology Monitor 14(1): 7.

Soskolne, Colin L. 1993b. "Introduction to Misconduct in Science and Scientific Duties." Journal of Exposure Analysis and Environmental Epidemiology 3 (suppl. 1): 245–252.

Soskolne, Colin L., ed. 1993c. Journal of Exposure Analysis and Environmental Epidemiology 3 (suppl. 1): 297–320. Special issue, "Questions from the Delegates and Answers by the Panelists Concerning 'Ethics and Law in Environmental Epidemiology.'"

Soskolne, Colin L., and Last, John M. 1993. "CMA Epidemiology Guidelines." Journal of Occupational Medicine 35(2): 97–98.

Stolley, Paul D. 1985. "Faith, Evidence, and the Epidemiologist." Journal of Public Health Policy 6(1): 37–42.

Susser, Mervyn. 1977. "Judgment and Causal Inference: Criteria in Epidemiologic Studies." American Journal of Epidemiology 105(1): 1–15.

Susser, Mervyn. 1985. "Epidemiology in the United States after World War II: The Evolution of Technique." Epidemiologic Reviews 7: 147–177.

Susser, Mervyn; Stein, Zena; and Kline, Jennie. 1978. "Ethics in Epidemiology." Annals of the American Academy of Political and Social Science 437: 128–141.

Teich, Albert H., and Frankel, Mark S. 1992. Good Science and Responsible Scientists: Meeting the Challenge of Fraud and Misconduct in Science. Washington, D.C.: American Association for the Advancement of Science.

Terris, Milton. 1979. "The Epidemiologic Tradition: The Wade Hampton Frost Lecture." Public Health Reports 94(3): 203–209.

Terris, Milton. 1987. "Epidemiology and the Public Health Movement." Journal of Public Health Policy 8(3): 315–329.

Vineis, Paolo, and Soskolne, Colin L. 1993. "Cancer Risk Assessment and Management: An Ethical Perspective." Journal of Occupational Medicine 35(9): 902–908.

Weed, Douglas L. 1997. "Underdetermination and Incommensurability in Contemporary Epidemiology." Kennedy Institute of Ethics Journal 7(2): 107–127.

Weed, Douglas L., and Trock, Bruce. 1988. "Interactions and Public Health Decisions." Journal of Clinical Epidemiology 41(2): 207–209.

Westrin, Claes-Goran. 1993. "Ethical, Legal, and Political Problems Affecting Epidemiology in European Countries." IRB 15(3): 6–8.

Westrin, Claes-Goran; Nilstun, Tore; Smedby, Bjorn; and Haglund, Bengt. 1992. "Epidemiology and Moral Philosophy." Journal of Medical Ethics 18(4): 193–196.