Nonconsensual Medical Experiments on Human Beings

copyright 1997 by Ronald B. Standler

These are some notes that I prepared during the Christmas vacation in 1996, after I read a number of cases and books about medical experiments on human beings. I begin with summaries of facts and my personal comments on several famous (notorious) cases, then I give my personal draft of a set of rules that might offer better protection to research subjects than the current system in the USA.

Table of Contents

Nürnberg Code

experimental horrors in USA after Nürnberg Code
  1. Tuskegee syphilis experiment
  2. Harold Blauer
  3. high oxygen to premature infants
  4. injections of cancer cells
  5. hepatitis in retarded children
  6. Cincinnati radiation experiments
my observations on these famous experiments

experiments on prisoners?

Standler's draft code

Improvements in medicine and surgery during the last two centuries have greatly extended the average lifetime of people and reduced the pain and disability following various injuries and diseases. Because medicine and surgery have a strong basis in scientific knowledge, it is absolutely necessary for medical progress that there be experiments on humans. There can be no doubt that human experimentation is necessary and desirable. Society owes a great debt to the many creative physicians and surgeons who have performed ethical experiments and advanced medical technology. If there be any doubt, I say that I am definitely in favor of research, but we should also insist on ethical conduct.

There is also a dark side to human experimentation: a long history of dangerous and harmful experiments performed on nonconsenting patients. Not only was informed consent not obtained, but the physician often fraudulently described the experimental procedure as either a diagnostic procedure or a treatment for the patient's condition, when the physician had no reason to believe that the patient might benefit from the experiment. History has shown that nonconsensual experiments are often performed on captive people in an institution, particularly people who society has regarded as "less worthy" (e.g., Jews in Nazi concentration camp, mentally retarded people in institutions, Negroes, indigent patients, ...). Such people are unable to decline or reject the experiment and few people will ever know what really happened.

Regulation of medical experimentation on human beings must consider the following:
  1. rational need of researchers to have access to human subjects. Experiments on human subjects are performed after in vitro experiments and after experiments on animals have shown that a drug or technique has a reasonable possibility of benefiting human beings. In assessing the desirability and acceptability of the experiment, one should consider the severity of the disorder (e.g., terminal illness vs. minor inconvenience) together with the possible side effects of the experimental treatment.
  2. potential subjects must give informed consent (i.e., consent after experimenter has made an honest, complete disclosure of risks). Informed consent is necessary for the personal autonomy of the subject.
  3. researcher's conflict of interest, in which researcher wants a large number of subjects to consent and then wants all of these subjects to complete the entire experimental program. However, the researcher, as a physician, also has an obligation to neither harm nor exploit the patients/subjects.
In my opinion, these three considerations are not to be balanced against each other, but are to be independently satisfied.

The conflict of interest is worth exploring further, as it is the key to analyzing much of the ethics of human experimentation. Patients expect safe, effective treatment for their condition and patients trust their physician to supply unbiased advice. But, in human experimentation, the relationship is often not physician to patient, but is researcher to subject. The researcher gets paid to do the clinical trial or experiment, and may also receive royalty income from new drugs, devices, or cell cultures that are developed in the experiment. The researcher may even become famous as the author of a landmark paper in the archival medical literature, which is a kind of immortality. In contrast, the subjects get pain and sometimes permanent injury, or even death, from the experiment.

A copy of the Nurnberg Code is posted at the U.S. Holocaust Memorial Museum's web site.

Condensed Nürnberg Code

1. Absolutely essential to obtain voluntary, informed consent of every human subject. Voluntary means without any element of force, fraud, deceit, duress, or coercion. Before asking his/her consent, must inform of all inconveniences and hazards reasonably to be expected and the effects upon his health or person that may come from his/her participation in the experiment.

2. Experiment must be designed to yield results for the good of society, unprocurable by other means of study.

3. Animal experimentation should precede experiments on humans.

4. Must avoid all unnecessary physical and mental suffering and injury.

5. Do not perform experiments in which there is reason to believe death or disabling injury will occur.

6. The degree of risk taken by subjects should never exceed that determined by the humanitarian importance of the problem to be solved by experiment. [Standler's comment: How do we apply this vague, overbroad principle?]

7. Proper preparations should be made and adequate facilities provided to protect the experimental subject against even remote possibilities of injury, disability, or death.

8. The experiment should be conducted only by scientifically qualified persons.

9. Human subject may withdraw consent at any time if he has reached the physical or mental state where continuation of the experiment seems to him to be impossible.

10. Scientist must terminate experiment at any time, if he has probable cause to believe that continuation of experiment is likely to result in injury, disability, or death to the experimental subject.

Standler's comment: Nr. 5 and Nr. 10 mean that we can not conduct experiments with new general anesthetics, nor can we do experimental major surgery, because of the risk of cardiac arrest. This code also appears to prohibit innovative treatments on dying patients. Results of such experiments could be useful to save lives of innocent people. As evidence that the writers of the Nürnberg Code overreacted, in Nr. 5 they forbade experiments that involved risk of death or disabling injury "except, perhaps ... the experimenting physicians also serve as subjects". It is not realistic for the physicians to be both subjects and also monitoring their condition. Furthermore, weasel words like "except, perhaps" have no place in a code of conduct. In Nr. 9, a subject should be able to withdraw consent at any time, for any reason, or no reason at all. I conclude that this Code is only a first draft, and not a particularly good one.

Society has no problem with asking military personnel to die for their country, why can't society tolerate people dying during medical research? People voluntarily engage in risky recreational activities, why can't they consent to potentially hazardous medical experiments?

The modern version of Nürnberg Code is the 1964 Declaration of Helsinki, by the World Medical Association. Various codes of ethics for physicians have been posted by the Center for the Study of Ethics in the Professions at Illinois Institute of Technology.

Notice that informed consent for medical experiment is different from consent to diagnostic and therapeutic procedures during regular health care. Before a medical experiment, the risks must always be disclosed to the subject, whereas there is often a therapeutic privilege to withhold full disclosure during regular health care.

Joseph Fletcher, in Experiments on Humans presented to the World Health Organization in March 1976, lists the following 15 factors to consider:
1. risk for both subjects and investigators
2. avoidance of harm
3. reportage (professional probity)
4. animal studies before tests on humans
5. consent (competence as to freedom and information)
6. benefit (proportionate good to be gained)
7. avoidance of deception of subjects
8. design of experiment to scientific standards
9. privacy, confidentiality
10. motivation of participants, both subjects and investigators
11. cost of experiment, part of risk-benefit calculation
12. fraud (deception in reporting)
13. in-course cancellation by subjects or investigators
14. monitoring
15. peer review, public review
Fletcher also mentions:
16. liability for injuries to subjects
17. guidelines for selection of both subjects and investigators

experimental horrors in USA after Nürnberg Code

The horrors of the so-called medical experiments in Nazi Germany were repeated many times, albeit on a smaller scale, by physicians in the USA. There were two provocative articles in New England Journal of Medicine [H. Beecher, 274 NEJM 1354 (1966); D.L. Rothman 317 NEJM 1195 (1987)] that suggested that an appreciable fraction of medical experiments in the USA failed to obtain informed consent. Unfortunately, neither author provided citations for the allegedly unethical studies, because "citing specific articles would distract attention from larger points". While that may be true, not providing citations means that readers can not verify the accuracy of what the author claims, so these two articles only give an unsupported opinion, and do not serve as a resource for scholars. One wonders if the author was really trying to avoid criticism from wounded colleagues, who had conducted unethical experiments and who now wanted to evade responsibility.

Keep in mind that the following are not horror stories about a few crazy, wayward physicians. This is mainstream medical research, some of it sponsored by the U.S. Government, some of it done by famous physicians. The following American experiments were after the Nürnberg Code was written.

Tuskegee syphilis experiment

The deliberate failure to treat a group of male Negroes in Macon County (near Tuskegee), Alabama who had syphilis begun in 1932 and ended, by unfavorable publicity, in 1972. This experiment is difficult to discuss, because so much was wrong with it. In my opinion, there are three major mistakes. Page citations are to Bad Blood by James H. Jones, a historian and scholar in bioethics.

  1. The U.S. Public Health Service (PHS) began using penicillin to treat syphilis in 1943 (p. 178) and penicillin became generally available in 1953 (p. 211). Despite this safe and effective treatment, which would have halted the progression of syphilis in the subjects, penicillin was deliberately withheld from the subjects. Not only did the PHS not give penicillin to their subjects, but the PHS (1) repeatedly distributed lists of names of subjects to local physicians and instructed the physicians not to give penicillin to these subjects (pp. 144-45, 162-63) and (2) supplied sham "treatments" (e.g., aspirin at p. 147, 160) to subjects, in an attempt to discourage subjects from seeking treatment elsewhere. The failure to treat this communicable disease violated Alabama law. (pp. 178, 212) In 1950-51, the PHS knew "that we have contributed to their ailments and shortened their lives." (p. 182) In the 1970's, the PHS offered the pretexual and lame excuse that penicillin would have harmed the participants if it had been given (pp. 8, 195), but when penicillin was given in 1973, there were no adverse reactions (p. 215). If, indeed, there were legitimate doubts about the safety and efficacy of penicillin in patients with long-term syphilis, administration of penicillin to these subjects would have made a worthy experiment in the late 1940's or early 1950's.
  2. Dr. Vonderlehr, of the PHS, practiced a fraud on the subjects by offering a painful lumbar puncture as a "special free treatment", when, in fact, the procedure was purely diagnostic and only for the benefit of the researchers.
  3. The subjects were not informed that they had syphilis (pp. 71-74, 219). Admittedly, informed consent would be extraordinarily difficult, because of the lack of education of the subjects: most were illiterate and many did not know their last name (pp. 4, 218). One of the best educated subjects had completed only eight years of school. To the extent that the subject's lack of education made informed consent impossible, these people should never have been allowed to suffer for the benefit of the physicians. It would have been a different matter if the subjects had been treated benevolently, for example, given penicillin prior to 1954.
To me, the most important question about the Tuskegee experiment is:

"Why was this experiment begun and continued?"

The project was begun in 1929 as a syphilis control program, using the standard therapy of 1929-33: a series of at least twenty IM injections of arsenic compounds, supplemented by topical applications of mercurial ointment. (pp. 45-90) However, there was not enough money for full treatment, despite the application of the PHS to private foundations. The subjects received only eight injections, which cured only 3% of them (p. 119).

There was a very high incidence of syphilis among Negroes in Macon County: 36%, which should be compared to the national average for Caucasians of only 0.4%. Given that these black men were living in distressing poverty (pp. 61-62, 83, 107, 201) and ignorance, it was easy to get them to follow orders from white physicians. When it was not financially possible to treat these subjects, the physicians may have looked for other things that they could do with this docile and relatively immobile group of people. In 1933, the long-term complications of syphilis were well known in Caucasians, but the common view amongst physicians was that Negroes responded differently to disease than Caucasians. So, apparently, the physicians decided to study the progression of untreated syphilis in Negroes, ending only when all of the subjects had been autopsied. I say "apparently", because with all of the bureaucratic obfuscation in the 1970's, it is not clear what the motivation really was in 1933.

I get the impression that the PHS saw these subjects as a wonderful opportunity for an experiment, when the PHS's first choice – a treatment program – was denied for lack of financial support, they found another experiment to do with this group of subjects. Support for my view is given by the continuation of the experiment in 1951: "the widespread use of ... [penicillin] had practically eliminated the possibility of finding another large group of syphilitic patients". (p. 179)

Given the unethical aspects of this experiment, why was it allowed to continue until stopped by public outcry in 1973? Apparently, the long-term experiment was never seriously questioned – as a matter of bureaucratic inertia – since it had continued for such a long time and was allegedly valuable. Further, when new managers were appointed, these new managers were reluctant to rescind decisions of their mentors. (p. 178, 180)

Bad Blood (at p. 190) alleges that the first physician objected to the experiment only in 1965. In 1966 a social worker at a PHS VD clinic in San Francisco objected to the study, and, in 1972, he tipped a reporter at Associated Press. (pp. 191-193, 203-205). The initial reaction of the PHS was that they had a public relations problem on their hands, but they had done nothing unethical. (p. 201) Remember, the Tuskegee experiment was no secret among physicians who worked on sexually-transmitted diseases: results from this experiment were published in medical journals. I think it remains an important and unanswered question of why the entire medical establishment was so blind to ethics for more than forty years.

other points of Tuskegee experiment

From the viewpoint of 1996, it is shocking that women were omitted from the study. But, by not treating the men in the study and by not informing the men of the nature of the disease, the PHS permitted many women to become infected with syphilis and many infants to acquire congenital syphilis. (pp. 104, 165, 215) Around 1930, 62% of the men in the study had congenital syphilis (p. 76), so the PHS simply allowed the disease to continue, nonetheless, one would expect better of the PHS.

A study of nearly 2000 untreated syphilitics in Oslo, Norway from 1891-1910 had been published and a follow-up study of them was published in 1929. (pp. 10, 92-3, 167, 183) Why did the PHS need to repeat the Oslo experiment? The common view amongst physicians in the 1930's was that Negroes responded differently to disease than Caucasians.

Some commentators mention that the PHS paid for the subjects' funeral expenses as a way of inducing consent to autopsy. I see nothing wrong with such compensation.

Finally, despite the suffering of the participants, it is uncertain how much valid, new medical knowledge was published as a result of this study. A 1933 review by the American Heart Association said the results of X-ray examinations of syphilitic hearts had "very little, if any, value". (p. 139) No one knew the exact number of subjects. (p. 181) Twelve of the two hundred controls acquired syphilis in the first six years of the study. (p. 176) Some of the subjects may have received penicillin for other infections. (p. 202) The autopsies were conducted until 1952 with "high-grade Neanderthal equipment". (p. 184) And, of course, these supposedly untreated subjects had received some treatment in the 1930's, which cured about 3% of them and had unknown effects on the progression of disease in the remainder. (pp. 119, 131, 173-76, 182, 202) I am also concerned that chronic malnutrition might induce pathologies of comparable severity to partially-treated syphilis. So, in the end, this ghastly experiment tells us nothing about untreated syphilis. And since penicillin is an effective cure for syphilis, we have no reason to know about the progression of untreated syphilis.

There was also little valid medical knowledge from the so-called experiments in Nazi concentration camps. This parallel raises the question "Are unethical studies also likely to be invalid?". While I acknowledge that technical skills in biology and statistics are entirely different from knowledge of morality and ethics, I nevertheless maintain that a competent scientist is zealous about both technical and ethical concerns. In particular, I don't see how a physician can work amongst suffering people without being aware of, and concerned about, their suffering. The moment that the physician sees subjects (as if they were a inanimate object), instead of patients or people, the physician has lost some of his humanity. If the physician can't see suffering, how much else can he not see?

At least 28, perhaps more than 100, of the 399 subjects had died from syphilis. (pp. 1-2) Litigation against the PHS, et al., was settled in Dec 1974, 18 months after suit was filed, for $ 37,500 to each surviving participant and $ 15,000 to each decedent's estate. (p. 217) As yet another piece of exploitation of the subjects, the attorney who brought the suit received 1/8 of the settlement, which gave approximately one million dollars to the attorney – quite a nice income in 1974 for a case that did not go to trial. (p. 217) Interestingly, this black attorney included neither any predominantly black institution nor any black physicians or nurses as defendants (pp. 216), although the Tuskegee Institute and a number of black physicians had been involved in the experiment, which was operated from the beginning until 1965 by one black nurse. While one might be reluctant to blame a black nurse for following orders of white physicians, especially during the 1930's and 1940's (pp. 151-169), "following orders" was not an acceptable excuse for German war criminals.


In Dec 1952, Harold Blauer, a civilian in excellent physical health, but depressed following his divorce, was voluntarily admitted to the New York State Psychiatric Institute for treatment of depression. While at the Institute, he was injected on five different occasions with three different mescaline derivatives supplied by the U.S. Army Chemical Corps to determine the clinical effects of chemical warfare agents in a research project that was classified secret. He reluctantly consented to the first injection, which was fraudulently offered to him as a treatment for his depression. After each of the first four injections, the subject told the nurses that he did not want any further injections, because of his adverse reactions. However, the nurses and staff motivated him to further participation by telling him that he would be returned to Bellevue Hospital if he did not continue with the injections. After the fifth injection, which was 16 times stronger than the previous injection of this new compound, his body stiffened, his teeth clenched, and he began frothing at the mouth for 2 hours until he lapsed into a coma and died. The death certificate attributed the death to "coronary arteriosclerosis; sudden death after intravenous injection of a mescaline derivative." The subject's ex-wife filed suit for medical malpractice against the State of New York, which operated the Institute. There were numerous conferences between the state's attorney, and attorneys for both the U.S. Army and U.S. DoJ. The state's attorney told the U.S. that he believed that, if all the facts were disclosed to plaintiff, the suit was worth at least $ 60,000. The Army then retrieved all of the medical records and took them out-of-state, so that they would not be discovered in the medical malpractice litigation. The state never supplied these records during discovery, despite a court order for production of documents. The malpractice claim was settled for $ 18,000 after false and misleading statements were made by the defense: namely that Blauer died because he allegedly had a weak heart and that the drugs were administered for diagnosis or therapy for Blauer's depression. The U.S. Government secretly reimbursed the State of New York for half of the settlement in 1955. The physicians who performed the experiment, the hospital, the State of New York, the U.S. Army Chemical Corps that sponsored the research and supplied the chemical for injection all conspired to conceal evidence or submit false documents. In Aug 1975, 23 y after the subject's death, the Army contacted Blauer's daughter and issued a press release about finding the case file in a safe. Blauer's ex-wife had died, so Blauer's daughter reopened the litigation, which resulted in a complicated series of cases.

Barrett v. State of New York, 378 N.Y.S.2d 946 (1976)
Barrett v. Hoffman, 521 F.Supp 307 (S.D.N.Y. 1981), rev'd 689 F.2d 324, cert. denied, 462 U.S. 1131, remanded 622 F.Supp. 574, aff'd 798 F.2d 565.
Barrett v. U.S., 660 F.Supp. 1291 (S.D.N.Y. 1987) (awarding $ 702,044 to daughter).
Barrett v. U.S., 668 F.Supp. 339 (S.D.N.Y. 1987), aff'd 853 F.2d 124, cert. denied 488 U.S. 1041
Barrett v. U.S., 1991 WL 60365 (attorneys billed $ 1.7 million in these cases)

A court in 1987 concluded that the LD50 studies in animals were "grossly inadequate" according to the standards in the early 1950's: the Army found that all mice died at 100 mg/kg and no mice died at 12 mg/kg, but the Army performed no further animal experiments. Further, before the judge approved the settlement in 1955, the state's attorney had an ex parte meeting with the judge, at which the attorney falsely represented that source of the drug was the Army Medical Corps, to give the impression that the experiment was therapeutic. The judge then required that the settlement be increased from $ 15,000 to $ 18,000.

high oxygen concentrations to premature infants

Premature infant in 1953 given high dose of oxygen as part of experiment, although physician who was responsible for infant's treatment (and aware of preliminary results that showed high oxygen caused blindness) had ordered oxygen "reduced ... as tolerated" and infant was doing fine. No attempt to obtain informed consent from parents. Researchers observed that infant's eyes swelled, but high oxygen was continued. At trial, plaintiff was totally blind and unable to find employment. Liability for both hospital and physician who ordered change to high oxygen.
Burton v. Brooklyn Doctors Hospital, 452 N.Y.S.2d 875 (1982)

injections of cancer cells

There were intradermal injections of live human cancer cells into 22 chronically ill, debilitated non-cancer patients in 1963 without their consent in the Jewish Chronic Disease Hospital case, to learn if foreign cancer cells would live longer in debilitated non-cancer patients than in patients debilitated by cancer. Lump at injection site disappeared approximately seven weeks after injection. Research funded by U.S.P.H.S. and American Cancer Society. The subjects were not told that the injection contained cancer cells, because the physicians "did not wish to stir up any unnecessary anxieties in the patients" who had "phobia and ignorance" about cancer. Physicians claim each patient gave "oral consent", but a material fact was not disclosed to patient and many of the patients were not in a physical or mental condition to give valid consent. Hospital administration tried to cover-up lack of consent, and some written consents were fraudulently obtained after the fact. Three physicians at the Hospital resigned when the administration did not seriously consider their complaints about the experiment. The chief of medicine at JCDH and the principal investigator were placed on probation for one year by the New York State medical licensing board, as a result of a unanimous guilty verdict on fraud/deceit and unprofessional conduct. Two years later, the American Cancer Society elected the principal investigator to be their Vice-President.

Hyman 251 N.Y.S.2d 818 (1964), rev'd 258 N.Y.S.2d 397 (1965) (litigation regarding whether director of hospital could have access to patients' medical records)
Katz, Experimentation with Human Beings (1972) pp. 9-65 recounts more than you want to know about this experiment, including affidavits of physicians on both sides.

In passing, it is noted that several hundred postoperative gynecology patients at Memorial Hospital had the same injections, also without consent, but there was no fuss about that. (Katz at 11, 23, 27, 37) Why was it necessary to inject human cancer cells into more than 300 healthy subjects; wouldn't a smaller number of subjects be adequate? There seems to be a genuine scientific controversy about whether the injections were dangerous: some cancer experts said that it was impossible to transplant human cancer cells from one person to another, as the donee's body would reject the foreign cells. But there is one documented case of transplantation of melanoma from daughter to mother (Katz at 309). And several physicians testified that the injected cells might produce cancer years later.

If a physician were to ask healthy people if they would willingly agree to have live cancer cells injected into them, the healthy people would certainly say "NO!" The practical result is that, if this research is to be performed on a large number of subjects, the physician performs his experiment on nonconsenting patients. This point raises a number of interesting questions. Is this experiment so important to society that the requirement of informed consent should be suspended? Is it really necessary to have a large number of subjects, or could we get adequate information from experiments on a few brave volunteers? I believe informed consent is always necessary when subjects are participating in an experiment that can not benefit them.

hepatitis in retarded children

Severely retarded children at the Willowbrook State Hospital in New York injected with hepatitis virus. This Hospital did not admit new patients after 1964, unless their parents "consented" to the experiment. Consent forms implied that children were to receive a vaccine against hepatitis, when the protection was actually from a hopefully "subclinical" infection. Physician made excuse: fecally-borne viral hepatitis was so prevalent at the Hospital (because approximately 70% of children had IQ below 20 and were not toilet trained) that children routinely became infected 6 to 12 months after admission. But isn't that like murderer saying that victim was going to die sometime, all he did was kill earlier and under better circumstances?

Cincinnati radiation experiments

Cancer patients (mostly Negroes of below-average intelligence who were charity patients) during 1960-72 in Cincinnati were exposed to large doses of whole body radiation as part of an experiment sponsored by the U.S. military. None of the subjects gave informed consent, they thought they were receiving treatment for their cancer. Subjects experienced nausea and vomiting from acute radiation sickness, pain from burns on their bodies, and some died prematurely as result of radiation exposure.

In re Cincinnati Radiation Litigation, 874 F.Supp. 796 (S.D.Ohio 1995)

On 5 May 1999, a settlement was announced in which the defendants paid US$ 5.4×106, but defendants admitted no wrongdoing.

continuing problems

In April 1999, all research projects at the Veteran's Administration West Los Angeles Medical Center were shutdown after many allegations of medical research performed on patients who did not consent. An investigation showed that not only was research being done on patients who had not given informed consent, but also that research was being done on patients who had expressly refused consent. Investigators found multiple violations of the government's code of research ethics. As is typical of government bureaucracies, the proposed solution was more management and more review, not criminal prosecution, and not termination of employment of unethical personnel.

my observations

I draw the following conclusions from these experiments.

  1. Physicians are often curious about details of how some disease or disorder progresses (e.g., syphilis, retrolental fibroplasia). They are not satisfied to know the cause and cure of the disease – they want to know the ghastly progression when the victims are untreated.
  2. Some physicians will resort to fraud or deception in order to obtain "consent"; other physicians look for helpless people in institutions (e.g., children, insane adults, prisoners, mentally retarded, chronically ill) to use as experimental material. The conclusion from this view is that some physicians see their patients as property under their control, not as independent lives.
    Do physicians choose incompetent people (e.g., insane, mentally retarded) because it is convenient to perform experiments on subjects who do not object or because these people are somehow less worthy (i.e., ok to harm dog or retarded human, but not normal human)? I can find no answer to this question in the medical literature – apparently, it is too unpleasant a question for physicians to consider.
  3. These physicians are absolutely convinced that what they are doing is justified in the name of scientific progress. They resent any implication that what they are doing is unethical.

The first point suggests that some physicians will continue to do experiments that society considers as "not worth doing", unless law regulates physicians. The second and third points suggest that some physicians are both powerful and arrogant. There is a clear need for non-physicians to review proposed medical research and restrain physicians, before they hurt their patients.

After the Tuskegee experiment, the U.S. Government issued 45 CFR § 46.101-117 to stop continuing atrocities from physicians in the USA, when federal money was used for salaries or at least some of the expenses. However, physicians doing research often regard ethical review as a mere bureaucratic inconvenience.

The number of unethical physicians implicated in the above experiments is very small, compared to the total number of physicians. However, the existence of a long series of atrocious examples suggests that many physicians may engage in informal experiments on some of their patients without any informed consent.

I suspect that very few physicians deliberately devise an unethical plan. My guess is that physicians are zealous about doing the experiment and they genuinely believe (perhaps correctly) that the result will be much more important to humanity than the pain, suffering, and possible death of a few experimental subjects. This kind of justification calls to mind the cliché, "You can't make omelets without breaking eggs." On careful, philosophical analysis, which the physicians do not do, this justification crumbles for several reasons. First, eggs are property that can legitimately be used as the owner wishes, but a human being has a fundamental right of autonomy and each patient has the right to decide what is done with his/her body. As a consequence of this principle, a nonconsensual touching is a battery, a legal wrong. Second, stressing the good results from medical research is simply an example of "the ends justifies the means", which is a bogus philosophical rational.

The practical problem of ensuring that human experimentation be done in an ethical way is extremely difficult. On one hand, there are already too many artificial burdens and impediments to medical research (e.g., lack of funding, inadequate time, obsolete facilities, members of review committees who have no understanding of research, repressive legislation). But we also can not ignore the fact that a number of innocent people in the USA have been harmed in the name of medical research and this abuse of patients must stop.

The solution is not to teach physicians about ethics. Everyone who enters medical school already knows not to perpetrate a fraud on patients and already knows not to regard "less worthy" people as proper subjects for experiments that are dangerous to them. The solution is to control zealous researchers who do not seriously consider ethics, because the experiment must be done at any cost. At the end of this essay, I suggest some specific rules to protect patients, while allowing research.

discuss experiments on prisoners

During wars, many young men are drafted, then put in harm's way. Society tolerated this dangerous involuntary servitude, because it was necessary for the greater good and a better future.

This same argument was used to justify medical experiments on criminals in prison [Rothman 317 N.E.J.M. 1195 (1987)] during World War II. Because a person is in prison does not relieve him of an opportunity to serve his country and human progress, by participating in medical research. "Every man must do his part."

If there currently is no draft and no military conflict to endanger normal people, we can still use this argument, because disease always threatens society.

It seems to me that nonconsensual medical experiments are wrong. The fact that young men are killed in wars to which they did not consent is a separate injustice, which is technically irrelevant to discussing medical experiments.

However, I think that it is permissible to conduct medical experiments on prisoners, if they volunteer and provide informed consent. Prisoners who are sane still have some rights of self-determination and autonomy; it is paternalistic to declare that no prisoner can consent to research. Prisoners have harmed society and participating in medical experiments is one of the ways that they can volunteer to help society, as a way of clearing their conscience. The prisoners should not receive a reduction in their sentence as the result of participation in medical experiments, since their participation is not part of their punishment, but an independent act. Further, the possibility of reduction of sentences would inject an element of coercion into the prisoner's decision. However, it is acceptable to pay the prisoners money for their participation. Yes, there are ethical concerns about allowing medical experiments on prisoners who volunteer: there is always a question of the sincerity of their voluntary decision, but this question is present with any subject of a medical experiment, as discussed in the next paragraph.

It is strange that liberals who object to allowing prisoners to voluntarily participate in medical research don't blink an eye at allowing seriously ill patients to participate in research. There can be no doubt that someone who is seriously ill is frightened, perhaps fearful of his/her life. By characterizing the experiment as a "new treatment", the experimenters can exploit the natural desire of the patient to receive the newest [and presumedly the best] treatment possible. Participating in an experiment is also a way to get admitted to a superior hospital or to be treated by outstanding physicians (e.g., professor of medicine). See Ingelfinger, 287 NEJM 466 (1972).

Just as most lawyers don't blink an eye at helping "guilty" criminals go unpunished, most physicians are not bothered by death. Professionals' insular view is not shared by the community. Most people have no experience making difficult decisions.


Technically, coercion requires that threat/harm must come from coercer. It is not coercion either to refuse to help someone or to put conditions on the help, provided that the person refusing or making conditions did not cause the victim's predicament. For example:
1. Poor student applies for educational loan from bank. Bank's conditions, or refusal to provide loan, is not coercion, because the bank caused neither student's poverty nor student's need for money.

2. Man drives woman out into desert and then gives woman a "choice": consent to sex or be abandoned in middle of desert. Man coerced woman, because man caused woman's plight. Coercion invalidates any "consent" that might be given.

Standler's draft code

Note: This draft has no legal significance. It is only a sketch of what I believe the law should be.

  1. experiments on animals first
    Animal experiments must be conducted before any human experiments. Animal experiments may demonstrate that new treatment is ineffective or dangerous (i.e., effective, but high incidence of death from treatment), and therefore human experiments may be avoided. Animal experiments may also disclose specific side effects that can be expected in human subjects, so that experimenters may monitor for these effects.

  2. only do new experiments
    The researchers must do a through search of the medical and biological literature before doing any human experiments. Work that has been previously reported should not be repeated, unless there is a significant improvement in methodology or reason to suspect fraud or mistake in previous work.

  3. design of experiment
    The expected benefit to society should outweigh the foreseeable risks to the subjects. The benefit to society should be judged by attorneys and clergy, not by the experimenters themselves, and not by other physicians, to get some objectivity. The approval of an experimental protocol should be made by a review committee whose members actually voting (including neither abstain ballots nor absent members) have a majority of nonphysicians.

    The experiment should be conducted in a scientific manner, with a double-blind fashion, with a control group. A competent mathematician with experience in both medical/biological experiments and statistics should review the experimental design.

    It is unethical to give a control group of people a placebo when the members of the control group are suffering from a disease for which there is a treatment or palliative that has been proven to be more effective than placebo. Beyond ethical issues, society would be better served by comparing a new drug with standard drug, so that physicians know the "best" treatment, and not merely have a collection of drugs that are known to be "more effective than placebo".
    ASIDE: Obviously, drug companies will resist such a proposal, since it is more likely that their new drug will be found to be "more effective than placebo" than found to be "better than the standard drug" for the condition. Drug companies are in the business of making money, and they can make much more money from a unique new drug that is protected by patent, than from selling a drug for which there are inexpensive generic competing versions available. Perhaps the resolution is to have a three-way experiment: compare the new drug against both placebo and the standard therapy. Allow the new drug to be marketed if it is more effective than placebo, because the new drug might be a rational choice in some patients for whom the standard therapy is undesirable.

  4. consent
    All facts known to the experimenters that would be relevant to a reasonable person's decision to participate, must be disclosed in advance of consent.
    1. Are there any anticipated side effects to subjects? Will any of the experiments or diagnostic procedures be painful?
    2. Is there any foreseen possibility of death or permanent injury?
    3. Do any of the researchers have a financial stake in the outcome? (e.g., royalties from patent, investment in manufacturer)

    If subject is a patient of a physician who is not participating in the experiment, then the experimenters shall also make a full disclosure to the attending physician, as a condition of taking his/her patient. This rule increases the chances that disclosure will be understood and that experiment is actually reasonable.

    The disclosures and consent must be in writing with two originals (one retained by subject, the other retained by experimenter), with subject's signature witnessed by at least one person employed by neither experimenters, financial sponsor of research, nor institution in which experiment takes place. Suitable witnesses may be subject's spouse, adult children of subject, officer at bank where subject has checking account, subject's attorney. Failure to obtain such written consent with an independent witness shall be considered as a presumption that informed consent was not obtained. Each subject shall have not sign a consent form until at least 24 hours after it was given to the subject, so that the subject can carefully consider the risks and benefits, and make a choice without influence by the experimenters.

    If the subject is incompetent (e.g., minor, insane, mentally retarded, intoxicated, unconscious, in severe pain, etc.), then the disclosure and consent form must be signed by either subject's legal guardian, subject's spouse, or agent named in subject's durable power of attorney for health care. If the first person approached by experiments to give consent for incompetent refuses consent, then the experimenters may not request consent from others (e.g., alternate agents).

    Children and incompetent adults (i.e., retarded, insane) should generally not be subjects of experiments, because they can not personally give valid consent. However, if children or incompetent adults need treatment for their condition and there is reason to believe that the experimental treatment will be at least as effective as the conventional therapy, then their guardian may be asked to consent.

    People who refuse to be subjects in an experiment are still entitled to receive conventional therapy for their disease. It is not ethical to motivate someone to "consent" by giving the person a Hobson's choice of experiment or no treatment. However, the experimenter does not need to provide the conventional therapy if there is another physician in the vicinity who is competent to provide conventional therapy.

    The consent must be obtained by a physician who is named as one of the investigators in both the proposal for funding and in the experimental protocol that was approved by the ethics committee. The job of obtaining consent may be delegated to neither a nurse, resident, intern, nor medical student.

  5. experiments on normal people
    It is ordinarily not acceptable to subject normal people (i.e., not suffering from disease that is subject of experiment) to experiments that have a risk of death or serious injury. However, there are exceptions to this rule, which are discussed below. Normal people should be used for physiological research, but not for research on the effects of drugs or therapies. Normal people don't take drugs, so there is no point in learning the effect of drugs on normal people.

    After informed consent, small numbers of normal people may be exposed to organisms that cause disease, to measure the response of their bodies to the disease and the initial progress of the disease. Once symptoms appear that make the patient uncomfortable, the patient must be given appropriate medical treatment (e.g., antibiotics, analgesics, antipyretics) to relieve suffering and prevent serious injury.

    It is not acceptable to cause injury to a subject just to study the response to the injury. The experimenter must have a proposed treatment for the injury and there must be objective reasons to believe that the proposed treatment will be better than any accepted treatment.

    It is acceptable to cause death or injury to small numbers (e.g., 3 or 4) of normal people when the results of the experiment are important to medical science and the knowledge can not be obtained in any other way. For such experiments, subjects should be paid an amount of money that is adequate to induce their consent. Will such payments coerce unemployed or disadvantaged people to consent? No, because the experimenters did not cause the unemployment or disadvantage, so coercion is not technically present.

    Should criminals in prison be all allowed to participate? Yes [Accord Bailey v. Lally, 481 F.Supp. 203 (D.Md. 1979)], provided that they are sane and their consent is voluntary. Prisoners should be compensated with money for their participation, but prisoners should not receive a reduction in their sentence as a result of their participation in medical experiments.

    Even if they are not able to give legally-binding consent, children and incompetent adults should have the right to refuse to participate in any experiment that is painful and is not a treatment for a disease that they have. Their right to refuse might be expressed as reluctance to go to the room where the experiment is conducted, by crying or screaming, or by bodily motions that indicate extreme discomfort.

  6. experiments on victims of disease or injury
    Treatments, including drugs, should be tested on a group of people who are suffering from the disease, injury, or condition for which the treatment is appropriate and potentially beneficial.

    A neutral observer should periodically, depending on frequency of observation of patients, monitor the treatment group in the experiment for both (1) any evidence of serious side effects and (2) any evidence of startling benefits. If the treatment group is experiencing a high number of serious side effects and the disease for which they are being treated is not serious, then the experiment should be terminated early. If the treatment group shows extraordinary benefits, then it is unethical to continue the experiment, thereby continuing to deny the benefits to the control group.

    It is acceptable to perform experiments having a risk of death or serious injury on patients who have an injury or disease that is likely to be fatal. However, the experiment must offer some hope of personal benefit to the subject. For example, one may try a new anti-neoplastic drug on a cancer patient. One may not put a cancer patient in a tub of ice water and measure how long it takes for the patient to become unconscious, for a research on survival in icy seas.

  7. take care of subjects after the trial
    After the experiment, the experimenters continue to have an obligation to disclose to the subject any risks that the subject experienced as a result of the partial participation in the experiment. Accord, Mink v. University of Chicago, 460 F.Supp. 713 (1978) (1000 pregnant women given diethylstilbestrol during 1950-52 to test its ability to prevent miscarriage, women not informed that they were part of experiment, risk of cancer in children known as early as 1971, but university made no attempt to notify subjects until late 1975, court approves battery theory)

    Should some adverse event occur to subject(s) during experiment, subject is then entitled to both medical and disability payments:
    a. free medical care for the remainder of his life for any disorder that has at least a 10% probability of being related to the experiment
    b. payment of loss wages that he could otherwise have reasonably earned (including financial support of spouse and children after subject's death), and
    c. funeral expenses.

    Should research that is not funded by the government be required to purchase insurance for subjects, or post a bond, to ensure that the subjects are adequately protected?

  8. publish results
    The results of the experiment, regardless of the result, should be published in an archival medical journal, so that the entire community can benefit from the knowledge gained in the experiment. If the tested drug or treatment is a failure, it is particularly important to publish the results, so that future experimenters can avoid repetition of the research and avoid suffering of more human subjects.

    If work remains unpublished three years after the end of collection of experimental data, then experimenters should pay each subject an additional $ 100/year until such work is published in a peer-reviewed archival journal. The amount of this payment may be increased by a judge/jury if there is liability for unethical conduct of the experimenters.

  9. subjects share in reward
    Human subjects have a right to share in financial rewards from experiment, because their participation is essential. Subjects should be paid for their time and inconvenience during the experiment, plus reimbursement of their expenses and lost wages.

    If cells removed from human subject are the basis for a profitable commercial product, then subject who is the original source of cells should be paid a royalty as a percentage of gross sales. In addition, human subject has the right to have his/her name attached to any cell line that results from cells removed from his/her body; subject must either consent to naming, or waive this right, in writing.

    Aside: I believe that Moore v. Regents of the University of California, 793 P.2d 479 (Calif. 1990) was decided wrongly: Moore suffered from painful bone marrow aspirations and repeated travel from Seattle to Los Angeles, so that his physicians could make hundreds of thousands of dollars from Moore's cells. Further, Moore's physicians deceived Moore about purpose of visits and procedures.

  10. withdrawal of consent
    A subject may withdraw consent to continued participation in the experiment at any time, for any reason, or no reason. The subject need not declare a reason to the experimenters. After the patient withdraws, the experimenters continue to have an obligation to disclose to the subject any risks that the subject experienced as a result of the partial participation in the experiment.

  11. may publish unethical experiments
    Sometimes physicians may (improperly) perform experiments without first obtaining informed consent or by making fraudulent statements to the subjects. The punishment for such lapses in ethics is civil liability for the experimenters. However, if there are any valid scientific results from the study, the work should be published in an archival, peer-reviewed journal. The publication should contain a brief note that the reviewers found the study to be unethically performed and the reasons for this judgment briefly recited, as an additional sanction against the authors. (If the authors fail to publish in a timely way, see Nr. 8 above.) There is no reason to have a second group of subjects suffer pain and inconvenience to repeat the research in an ethical way. Indeed, the deterrent value to physicians who are tempted to be unethical may be greater if they are known to the profession as having conducted an unethical experiment. Further, the publication in a medical journal serves as a reminder to physicians who are designing experiments, unlike mention in a court proceeding, law journal, or newspaper.

  12. monitor experiment
    An independent physician should be appointed to monitor the experiment and stop the experiment if the treatment group has either (a) high incidence of side effects that are intolerable when compared to the potential benefit or (b) the treatment group is receiving a benefit that is extraordinarily better than the control group, so it would be unethical to continue the experiment and deny the benefit to the members of the control group. "Independent" means that the monitoring physician's name does not appear in the list of authors of the published research and the payment to the monitoring physician is the same, regardless of the results of the experiment.

posted on the Internet at:
created Dec 1996, minor revision 8 May 1999, links revised 18 June 2000

return to my homepage