Between December 1946 and August 1947, an American military tribunal at Nuremberg tried twenty-three Germans for crimes committed in the name of medicine. Most of the defendants at the Doctors' Trial were physicians: twenty of the twenty-three. They had plunged prisoners into icy water to study hypothermia, sealed them in low-pressure chambers that mimicked high altitude, rubbed bacteria and ground glass into deliberate wounds, and helped run the killing of disabled people judged unworthy of life. Most victims were camp prisoners with no say at all. Sixteen defendants were convicted and seven were hanged. This course begins with what medicine learned from this smaller trial, held alongside the main one.

The Nuremberg Code

The judges' verdict set out ten principles for permissible experiments on people, now called the Nuremberg Code. The first and longest says that the voluntary consent of the human subject is absolutely essential. The person must have the legal capacity to consent, must be free of force, fraud and pressure, and must understand enough to make an informed choice. The other nine points ask that an experiment aim at results useful to society that cannot be gained another way, rest on animal studies, avoid needless suffering, keep risks in proportion to the question, and let the subject stop at any time.

The Code was written by judges, not doctors, and no country adopted it as law. Many researchers outside Germany read it as a rule for barbarians that had nothing to do with them. That assumption proved costly.

Helsinki and the problem at home

In 1964 the World Medical Association, a federation of national doctors' groups, issued the Declaration of Helsinki. It separated research that might help the patient from research done purely for knowledge, and it allowed a legal guardian to consent for someone who could not. Its 1975 revision added review of research plans by an independent committee, a safeguard the Nuremberg Code lacked.

Meanwhile, abuse continued in democracies. In 1966 the Harvard anesthesiologist Henry Beecher published an article showing that respected researchers had run risky studies on people without proper consent, and leading journals had printed them. He described twenty-two examples. When the Tuskegee syphilis study was exposed in 1972, the United States Congress responded with the National Research Act of 1974, which created a national commission on research ethics and led to rules requiring institutional review boards, or IRBs, to approve federally funded studies.

The Belmont principles

The commission's most lasting product was the Belmont Report, published in 1979 and named after the Maryland conference center where it was drafted. It named three principles. Respect for persons means treating people as able to run their own lives and giving extra protection to those whose ability to choose is limited, such as children. Beneficence means not harming people and making the benefits of research as large, and its risks as small, as possible. Justice asks who bears the burdens of research and who receives its benefits. The report noted that poor patients on public hospital wards had long carried the risks of research while the resulting treatments flowed to private patients who could pay.

Each principle turned into a practical rule: respect for persons into informed consent, beneficence into a careful weighing of risks and benefits, justice into fair selection of who is asked to take part. In 1991 most US federal agencies adopted these rules in a shared regulation called the Common Rule.

Four principles and their critics

Also in 1979, Tom Beauchamp and James Childress published Principles of Biomedical Ethics, which offered four principles for clinical care as well as research: respect for autonomy, nonmaleficence, the duty not to cause harm, beneficence and justice. This approach, often called principlism, became the common language of hospital ethics committees. Its defenders say people of very different traditions can agree on these principles while disagreeing about ultimate questions.

Critics reply that four principles with no rule for ranking them can justify almost anything: when autonomy and beneficence clash, the framework itself does not say which wins. Albert Jonsen and Stephen Toulmin revived an older method called casuistry, which works by reasoning from clear paradigm cases to new and harder ones. Care ethicists add that principles drawn from contracts between strangers miss the relationships, such as parent and child or nurse and patient, in which most medical decisions are made.

Worked example: an asthma trial

A drug company wants to test a new asthma inhaler in teenagers aged twelve to seventeen. It proposes to recruit at a busy city clinic whose patients mostly come from low-income families. The hospital's IRB must decide whether to approve.

Respect for persons requires the permission of a parent and the agreement, or assent, of each teenager, with the study explained in plain language. The IRB worries that families may feel they must say yes to keep their own doctor happy, so it asks that someone outside the treating team handle recruitment. Beneficence requires that every teenager keep a rescue inhaler, that a safety board review results as they come in, and that the trial stop early if the new drug performs badly. Justice raises the hardest question. If the drug works, will families at this clinic be able to afford it, or will it be priced for the insured? The IRB asks the company to recruit at a suburban clinic as well and to explain its plans for access after the trial. Each step descends directly from what the Doctors' Trial revealed.