A key document that has helped define what's ethical in medical research for the last four decades has been revised to say that using placebos is unethical in nearly all experiments involving diseases that already have good treatments.
Representatives of the medical profession from 45 countries met on 08 October 2000 in Edinburgh, Scotland, and approved changes to the "Declaration of Helsinki," to say that any new treatments under study should be tested against "best current" treatments. Placebos can be used, however, where "no proven [treatment] exists."
Placebos are considered essential tools in distinguishing the effective from the ineffective in the flood of drugs and vaccines tested annually in human volunteers. The Food and Drug Administration requires their use in many studies, as does the agency's counterparts in other countries.
Taken literally, the new language would push hundreds, if not thousands, of clinical trials here and abroad beyond the boundaries of ethical acceptability.
The Declaration of Helsinki has no legal authority in the United States. However, as one of the handful of postwar documents designed to set international standards of behavior, its moral influence over national governments, ethical review boards and individual researchers is considerable. What practical effect--if any--the revision may have on medical research was not clear yesterday.
"I believe that it would be a mistake to rule out the use of placebos in well-designed research," said Greg Koski, director of the office for human research protections at the U.S. Department of Health and Human Services. "The Declaration of Helsinki has been viewed as a very important set of guiding principles for the ethical conduct of research. . . . I don't believe we would literally take the declaration as the literal basis for new regulations."
Margaret Dotzel, the FDA's associate commissioner for policy, said the agency would "have to look both at the language [of the revisions] and study it and make a determination how we would react to it."
The Declaration of Helsinki was produced in 1964 by the World Medical Association, which comprises representatives from the national medical societies of about 80 countries. The American Medical Association is the U.S. delegate. This is the fifth revision.
The latest one occurred, in part, because of controversy that arose three years ago over the use of placebos in studies seeking to find a cheap and easy way to decrease transmission of the AIDS virus from pregnant women to their babies. Conducted in Ivory Coast, Uganda and Thailand, the experiments gave infected pregnant women (and in some cases, their newborns) either placebos or short courses of the antiviral drug AZT near delivery.
The campaign for the new declaration language was spurred mostly by a desire to prohibit the use of placebos in studies in poor countries if their use in the same studies in rich countries would be considered unethical. The effect of the revision, however, is to greatly restrict the use of placebos everywhere.
"We say almost explicitly . . . that if there is treatment, then you cannot give a sugar pill to the control group," Delon Human, a South African physician and secretary general of the WMA, said yesterday.
The declaration previously had a paragraph about placebo use whose meaning was unclear. Some researchers believed it said that every person in a clinical trial--the control groups and the group getting the drug being tested--had to have the best treatment. Others believed it meant what the revised declaration now says clearly.
The document also contained several other phrases and ideas that were either unclear or being openly violated by the research community because they were out of date with current scientific methods and thinking. Whether researchers will be able to violate provisions of the newer, clearer and more high-profile document with the same impunity is uncertain.
Placebos are always used in clinical experiments where no alternative to the thing being tested exists. (New vaccines against diseases for which there aren't any vaccines are always tested against placebos, for example.) One of the many advantages of placebos is that they give every study participant the benefit of taking something he hopes will work--and that benefit is often considerable.
But placebos are also used in studies involving diseases for which there are good treatments. Most new anti-hypertensive drugs are tested in studies in which people with high blood pressure may get either the new drug or a placebo. Most new antidepressant drugs are tried out in groups of depressed patients, who may be assigned to get either the new drug or a placebo. Painkillers, anti-allergy drugs and other substances designed to relieve relatively mild symptoms are almost always tested against placebos.
The rationale for using them in those circumstances is that they help produce clear-cut results (and prevent the market from being flooded with drugs whose effectiveness is marginal or uncertain). For mathematical reasons, it is far easier to detect a drug's effect--and it requires fewer patients--when a new drug is tested against a placebo than when a new and an old drug go head-to-head.
Such studies have traditionally been considered ethical because the period in which a person would get a placebo is short and the risk of serious harm is minimal. In addition, all participants are volunteers, and may be willing to accept the small risk of taking a placebo.
Placebos are never used in studies of conditions--such as pneumonia or meningitis--in which the risk of death or other serious complication is likely to be high.
In the case of the AZT studies in developing countries, two physicians at Public Citizens Health Research Group argued that instead of a placebo, the women in the "control" group should have gotten the three-month-long AZT course that an experiment in the United States had shown was able to cut mother-to-child transmission by two-thirds.
Defenders of the placebo-controlled studies, however, said the key question was whether short-course AZT was better than nothing, which is what the African and Thai women would normally get, and not whether it was as good as long-course AZT, which none would have gotten. They also argued that possible toxicity of AZT in the Third World women (who had other health problems) could be evaluated only if some of them weren't taking the drug. Placebo-controlled trials were also cheaper and faster to do, potentially speeding the time when a useful prevention strategy was available in poor countries.