The first international code of ethics for research involving human subjects, the Nuremberg Code, was developed from 1947 to 1949 by the Nuremberg Military Tribunals during their trial of the Nazi physician-researchers; the Code's immediate purpose was to provide a set of standards for judgment of outrages committed in the name of science by the Nazi physician-researchers [3].

In 1964, the World Medical Association issued the Declaration of Helsinki [3, 4], which adapted the principles of the Nuremberg Code to the existential realities of medical research [3]. For example, the Declaration replaced the Nuremberg Code's first principle ("The voluntary consent of the human subject is absolutely essential") [4] with a recognition of the legitimacy of proxy consent for research involving children and persons with cognitive impairment. Unlike the Nuremberg Code, the Declaration of Helsinki reflects the fact that research protocols might, and often do, include components expected to provide direct therapeutic, diagnostic, or prophylactic benefit to individual subjects and that ethical justification of such beneficial modalities should differ from that of nonbeneficial procedures.

The 1982 CIOMS-WHO Guidelines added, among other things, a requirement for review and approval of all proposed research by an "ethical review committee" [2]. The Declaration of Helsinki, even in its third and most recent revision in 1989 [5], calls only for "consideration, comment and guidance" by "a specially appointed (independent) committee".

Until very recently, all international ethical codes and national laws and regulations were based on the assumption that medical research was hazardous to and exploitative of the subjects; this assumption reflects their origins as responses to the atrocities committed by Nazi physician-researchers, the calamitous experience with thalidomide, and, in the United States, exposes of such scandals as the Tuskegee Syphilis Study [5]. Since the mid-1980s, this assumption has been largely displaced by the equally incorrect perception of participation in research and access to investigational drugs as benign and beneficial; this shift was due primarily to the successful efforts of highly articulate and influential AIDS activists [5]. The writers of the 1993 Guidelines, recognizing the need for a balanced perspective, sought to encourage investigators to conduct ethical and beneficial research while maintaining necessary vigilance to safeguard the rights and welfare of research subjects [5].

When research is carried out in one country by investigators from another country, whose ethics should apply? This question, the most difficult one that confronted the writers of the 1993 Guidelines, is subsidiary to the larger question of whether ethics are universal. Ethical universalists hold that ethical standards, properly understood, are and ought to be the same everywhere [6, 7]. Ethical pluralists, by contrast, maintain that because systems of ethics are socially constructed within particular cultures and necessarily reflect their histories and traditions, variations across cultures are both expected and legitimate [8-10]. The 1993 Guidelines reflect a compromise. Although some ethical principles, such as respect for persons, beneficence, and distributive justice, are considered universal, the legitimacy of ethical pluralism within specified limits is acknowledged.

The 1993 document is 52 pages long and consists of 15 statements of guidelines, each of which is followed by a detailed commentary on how the guideline should be applied in specific contexts. Novel features can be found in several categories, of which three are informed consent, ethical review, and obligations of sponsors and host countries.

The Guidelines state that in some cultures, when investigators cannot make prospective subjects sufficiently aware of the implications of participation (in research) to give adequately informed consent, the decision ... should be elicited through a reliable intermediary such as a trusted community leader. In some cases other mechanisms may be more suitable. However consent is obtained, all prospective subjects must be clearly told that their participation is entirely voluntary, and that they are free to refuse to participate or to withdraw ... at any time without loss of any entitlement [1].

Each individual must receive all information that would be conveyed if the study were to be conducted in a developed country; "otherwise, assurance of freedom to refuse or withdraw would be meaningless" [1]. Furthermore, "In some cultures women's rights to exercise self-determination are not acknowledged. In such cases, women should not normally be involved in research for which societies that recognize these rights require informed consent" [1]. The Guidelines make provisions for affording women in such cultures access to investigational drugs and other therapies for which there are no equal or superior alternatives, even though they cannot give formal consent.

When research is initiated or sponsored by agencies in a developed country and carried out in a developing country, "the ethical standards applied should be no less exacting than they would be in the case of research carried out in (the initiating) country" [1]. Committees in the initiating country are assigned special responsibility for determining that

the scientific methods are sound and suitable for the aims of the research, (that) the drugs, vaccines or devices to be studied meet adequate standards of safety, (that) there is sound justification for conducting the research in the host country rather than in the country of the external sponsoring agency, and that the proposed research does not in principle violate the ethical standards of the external sponsoring country .... Committees in the host country have the special responsibility to determine whether the goals of research are responsive to the health needs and priorities of the host country. Moreover, because of their better understanding of the culture in which the research is proposed to be carried out, they have special responsibility for assuring the equitable selection of subjects and the acceptability of plans to obtain informed consent, to respect privacy, to maintain confidentiality, and to offer benefits that will not be considered excessive inducements to consent.

In short, ethical review in the external sponsoring country may be limited to ensuring compliance with broadly stated ethical standards, on the understanding that ethical committees in the host country will have greater competence in reviewing the detailed plans for compliance in view of their better understanding of the cultural and moral values of the population in which the research is proposed to be conducted [1].

The obligations of external sponsors of research to be conducted in developing countries are, in general, presented as prima facie obligations; that is, they may not apply in specific cases if competent authorities in both countries agree not to invoke them. For instance, research designed to develop a product (for example, a drug or vaccine) should be conducted only in host countries in which the disease for which the product is indicated is an important problem; any such product should be made reasonably available to inhabitants of the host country or community at the completion of successful testing. To help develop the capacity of the host country to carry out similar projects independently, sponsors are expected to employ and, if necessary, train local persons to function as investigators, research assistants, or data managers or to serve in other similar capacities. Sponsors are further expected, when appropriate, to make necessary health care services available to the population from which research subjects are to be recruited and to provide reasonable amounts of financial, educational, and other assistance to enable the host country to form independent and competent scientific and ethical review committees.

International declarations and guidelines issued by such agencies as the World Medical Association or CIOMSWHO do not have the force of law. The extent to which they have any effect on the conduct of biomedical research involving human subjects depends entirely on the influence they have on the development of national regulations and the policies of professional organizations and agencies that provide funding for biomedical research. United States federal regulations for the protection of human research subjects have been influenced substantially by the Nuremberg Code and the Declaration of Helsinki. According to a survey conducted by CIOMS, the 1982 CIOMS-WHO Guidelines have had a major influence on research policies and practices around the world [1]. The 1982 Guidelines had no perceptible effect on U.S. federal regulations because they were largely compatible with the regulations and because they were issued after the last major revision of the federal regulations was completed in 1981. The 1993 CIOMS-WHO Guidelines similarly do not create a need to consider substantial revision of the U.S. federal regulations; rather, they call attention to points that ought to be considered in addition to those covered in federal regulations—particularly when research is carried out in a developing country by investigators from a developed country.

Are these guidelines to be considered the final word on the ethics of multinational research? As Co-chair of the Steering Committee that supervised their development, I can say most emphatically that they are not. Ethical codes, guidelines, and regulations, if they are to remain vital and valid, require constant interpretation, reinterpretation, and occasional revision in the light of practical experience in the field.