ETHICAL, LEGAL, AND SOCIAL ASPECTS OF MEDICAL CARE

Session IV: Research

September 15, 1998

Prioritized Learning Objective:

1. To appreciate contemporary issues in research ethics.

2. To appreciate the importance of respect for persons, doing good and avoiding harm, and justice when conducting research.

3. To appreciate the concept of inherent coercion in human experimental subjects.

4. To be familiar with the Tuskegee, Sloan-Kettering and Willowbrook experiments.

5. To consider whether certain kinds of experiments related to weaponry should be conducted.

6. To appreciate potential ethical dilemmas in the use of soldiers as research subjects due to exceptional harm, military necessity, soldiers' willingness to make sacrifices in combat, and soldiers' inherently coercive situations.

7. To appreciate arguments for and against giving compounds not fully tested to servicepersons in the Persian Gulf.

8. To appreciate difficulties in arguing for or against the use of animals in research on the basis of their being a different species.

9. To identify and discuss ethical aspects of options to take when encountering animals that appear to be exposed to poor conditions.

10. To discuss ethical aspects of using animals for combat purposes.

Computerized Learning Experience: Lesson A

Required Reading:

1. Beauchamp, TL, Childress, JF: Principles of Biomedical Ethics, Oxford Press, 4th Edition, 441-447, 1994.

2. Howe, EG, Martin, ED: Treating the Troops. Hastings Center Report, 21-24, March-April 1991.

3. Vogel, S:In Baltic War Game, U.S. Navy Pampers a Different Breed of Sailor. International Herald Tribune, July 28, 1998, 10.

Prioritized Recommended Readings on Reserve in the Library:

1. Katz, J: The Regulation of Human Experimentation in the United States--A Personal Odyssey: IRB Vol. 9, Jan/Feb 1987.

   (An "insider's" historical view of the changes in research ethics in this country over the past several decades).

2. Jonas H: Philosophical Reflections on Experimenting with Human Subjects. From Experimentation with Human Subjects by Paul A. Freund ed. 1969, 1970 by the American Academy of Arts and Sciences.

   (A classic work stressing the importance of the value of respecting persons when doing research).

3. BELMONT REPORT: Ethical Principles and Guidelines for Research Involving Human Subjects.

   (A definitive ethical statement of the values which must be prioritized when research is conducted in this country).

4. Marshall E: Does the Moral Philosophy of the Belmont Report Rest on a Mistake? IRB Vol. 8, Nov/Dec 1986.

   (A dense, but thoughtful critique of the Belmont Report suggesting that it represents too many compromises and gives too little value to the need for protecting human subjects).

5. Veatch RM: The Patient as Partner: Ethics in Clinical Oncology Research. Johns Hopkins Medical Journal 151-161, 1982.

   (A thoughtful article expressing creative ways in which researchers could give research subjects more choices when they participate in experiments so that their own interests are better met and that they serve more actively as researchers' partners).

6. Vanderpool HY, Weiss GB and Appelbaum PS, et al: False Data & the Therapeutic Misconception: Two Urgent Problems in Research Ethics. Hastings Center Report, April 1987.

   (A "must read" article for researchers, discussing the tendency of researchers to sacrifice the scientific validity of their research to benefit patients and patients' tendency to believe that they will benefit from being subjects of research).

7. Kolata: Drug Found That Helps Heart Attack Survivors. Science 214:774-775, Nov. 13, 1981.

   (A discussion of the need to terminate research because of the benefits of propranolol for patients who had myocardial infarction).

8. Mahler: When to Obtain Informed Consent in Behavioral Research: A Study of Mother-Infant Bonding. IRB 8(3):7-9, 1986.

   (A state-of-the-art analysis of how best to study the effects of bonding, ethically).

9. Baumrind D: IRBs and Social Science Research: The Costs of Deception. IRB 1(6):1--4, October 1979.

   (The author articulates many psychological costs of using deception in research including some that are not so obvious).

10. Soble A: Deception in Social Science Research: Is Informed Consent Possible? Hastings Center Report 8(5):40-46, October 1978.

    (Soble proposes in this article that in addition to subjects giving prior general consent to being subjects in research involving deception (an approach introduced by Milgram) as an additional safeguard a friend or relative also serve as a proxy decision maker).

11. Rosebury: Medical Ethics and Biological Warfare. Perspectives in Biology and Medicine, pp. 512-523, Summer 1963.

    (A classic article on this topic).

12. Krickus J: On the Morality of Chemical/Biological War. J. Conflict Resolution 9(2):200-210, 1965.

    (More discussion of this topic, raising some distinctions between chemical and biological weaponry).

13. Caplan AL: Beastly Conduct: Ethical Issues in Animal Experimentation. Annals New York Academy of Sciences.

    (An overview of the major ethical arguments for and against the use of animals in research. The author believes that reasons given for using animals, in general, should be compelling).

14. Nelson JL: Animals, Handicapped Children and the Tragedy of Marginal Cases. Journal of Medical Ethics 14:1988, 191-193.

    (An article which philosophically explores the ethical bases and implications of using "marginal" humans such as those who are severely retarded in place of animals as research subjects).

15. Levine RJ: Uncertainty in Clinical Research. Law, Medicine & Health Care 16(3-4):1988, 174-182.

    (A review of the current debate in regard to the use of randomized clinical trials and how this information can best be communicated to patients).

16. Jackson, CM: The Fiery Fight for Animal Rights. Siegel, S: Grassroots Opposition to Animal Exploitation. Singer, P: To Do or Not to Do? Hastings Center Report, Nov/Dec 1989, pp. 37-45.

    (Several articles addressing the justifications of political action by groups supporting animal rights).

17. Ethical Theory and the Moral Status of Animals. Hastings Center Report, May/June 1990, pp. 4-30.

    (Several articles discussing the current status of ethical thinking and practice in regard to the use of animals in research).

18. Frisina, ME: The Offensive-Defensive Distinction in Military Biological Research. Hastings Center Report, May/June 1990, pp. 19-22.

    (A forceful analysis of the argument given against researchers conducting experiments that could be used for offensive or defense weaponry).

19. Smith, RJ: U.S. to Keep Producing Poison Gas. The Washington Post, October 9, 1989, A1, A8.

    (A discussion of President Bushes' decision in 1989 to continue to produce chemical weapons).

20. Donnelley, S: The Troubled Middle in Medias Res. Hastings Center Report, May/June 1990. pp. 2-4 (Special Supplement).

    (An article providing an overview of the animal rights debate which stresses the importance of respectful discussion).

21. Okie, S: Researchers Assailed for AIDS Studies on Pregnant Women in Third World." The Washington Post, September 18,1997, A13

    (An article depicting the recent controversy over whether to do research in Africa which would allow using a different standard by giving some research subjects with HIV infection a placebo.)

22. Editorial: "Pragmatism in Codes of Research Ethics."

    The Lancet 351 (90-98), January 24, 1998.

    (The writer argues that previous international standards should be abandoned to benefit citizens of other countries.)

23. "New law Means More Children to Serve as Research Subjects", Human Research Report 13 (2): 1-2, February 1998.

    (A brief report of a proposal to make it easier to do research on children.)

INTRODUCTION TO SESSION IV

September 15, 1998

During the first twenty minutes of this session, Dr. Moshe Shmuklarsky, who conducts research at Walter Reed Army Institute of Research (WRAIR), will present current ethical dilemmas encountered in researching agents to prevent servicepersons from developing malaria. During the last part of this session, Jon Spelman will present the drama, "Frankenstein."

In this session you will consider three basic issues related to research: ethical principles using human subjects in therapeutic and non-therapeutic research, military physicians' obligations when conducting research relevant to military weaponry and to use compounds during war that appear efficacious even though they have not been fully tested, and ethical considerations in regard to the use of animals in research.

Moral sensitivity in regard to research on humans has changed over recent decades. Until the notorious Tuskegee, Sloane-Kettering, and Willowbrook experiments,and those performed by Nazi physicians during World War II, the public tended to trust the medical profession and there were few formal procedures required to protect human subjects.

The Tuskegee experiments involved investigators' continuing to follow the course of syphilis in black men in Alabama even after treatments for syphilis had been discovered.

The Sloane-Kettering experiments involved physicians injecting live cancer cells into patients without informing them. Though the harm done to these patients seems to have been negligible, the failure to gain their consent was objectionable. The medical residents who performed these injections had to go over the head of the senior physician/investigator to bring this abuse to a halt.

The Willowbrook experiments used mentally retarded children institutionalized at Willowbrook, New York as subjects. The children were given hepatitis artificially in the hope of providing knowledge that might lead to an effective vaccine. This use of these subjects was then considered justified on two counts: first, those children who participated in the experiment would benefit because they would be less vulnerable to acquiring more severe hepatitis from natural causes; and second, since these children lived in unpleasant surroundings, those who participated in the experiment would benefit by residing in nicer quarters.

The ethical error raised in criticizing both justifications is that these researchers were exploiting these subjects' vulnerable predicament. This violated human dignity and also might have had the consequence of making it more likely that these circumstances would continue to be tolerated as opposed to being improved or eliminated. Subsequently, this institution has been closed and these children now live with foster parents in small groups.

Human dignity may also be violated in other inherently coercive situations. In the Kaimowitz Case, a prisoner who had been convicted of rape was asked to participate in an experiment in which he would be given brain surgery in the hope of reducing his impulses to commit violent sexual acts. The court refused to permit this research in spite of the prisoner's consent, because the surgery would interfere with his free thought and because he was a prisoner who would hope that this operation would result in obtaining an earlier parole. The court held that he could not freely consent because his consent was not wholly voluntary.

Inherent coercion extends to other groups, such as those who are poor. Suppose, for example, that large sums of money are offered to a person who is poor. This exploits poverty and is unethical because it treats persons unequally, using one group of human beings, the poor, as a means to other persons' ends as opposed to treating all persons as ends in themselves. Federal institutions now place strict limits on the amount of money that research subjects can be paid in order to reduce the potential for exploitation of this population.

Most experimentation permits some exploitation of persons' circumstances. Students wishing to be accepted into medical school sometimes participate as subjects of medical research, for example, to have a better chance of being accepted at medical school. Small degrees of exploitation are generally accepted as a necessary price for achieving research ends. If research subjects consisted only of those who volunteered for purely altruistic reasons, little research might take place. Even then, some "altruistic" volunteers might participate for "neurotic" reasons such as to punish themselves for wrongdoing, imagined or real, in their past. To avoid exploiting these research subjects, they might have to be precluded from serving as research subjects. How might this be done? What if you asked subjects questions concerning why they were volunteering? What if you required that they be screened by psychological testing or psychiatric interview?

Patients who are sick and participate in therapeutic research may also be exceptionally vulnerable since their consent may be motivated by the wish to get well. If the physician treating the patient is also the researcher, the patients' capacity to consent freely may be doubly compromised. In addition to participating because they want to get well, they may also fear that by not participating they could offend their physician and that their physician might unconsciously, if not consciously, provide inferior treatment. Studies such as those described in the sixth prioritized recommended reading suggest that patients--and some physicians usually believe that patients will benefit more by being a research subject than by being treated clinically? Why, theoretically at least, might they be wrong?

Recently, the most controversial issues involving human research have to do not with inherently coercive situations of patients due to their vulnerability, but with the conditions under which patients with HIV infection should have access to drugs which could be therapeutic. The two most important ethical questions are when randomized clinical trials are necessary and when the absence of a randomized clinical trial on a new drug should preclude persons with HIV infection from having access to that drug. Sam Kazman, an attorney, wrote in the Washington Post on July 16th, 1989, for example, "The inability of people with AIDS to obtain access to experimental drugs outside of government-approved channels has been a gross, and needless, compounding by the law of their medical agony....Why doesn't the FDA no longer stand between AIDS sufferers and the medicines they seek, so that the people with AIDS are free to obtain what they can from doctors, drug companies, and researchers...."

I will discuss both these questions later, but first I will discuss considerations pertaining to research using healthy subjects.

Healthy subjects are often selected to determine new drugs' toxicity and optimal dosages. The potential harm to healthy subjects usually is greater than the benefit. These subjects do benefit, however, from having the opportunity to be altruistic. I am reminded of persons who donate bone marrow regularly and are impatient during the period they have to wait before they can give again. Other benefits are more subtle. Servicepersons may participate in research during peacetime, for instance, to escape boredom. Subjects' boredom, on the other hand, may make their situation inherently coercive and this may warrant providing these subjects greater protection. The risks healthy persons are permitted to undergo in most instances are, relatively speaking, minor. Studies at the Walter Reed Army Institute of Research have rarely required subjects to undergo more harm than exposing themselves to several mosquito bites or, at most, a few days of high fever, headaches, muscle pain, and chills in research on new vaccines to prevent malaria. A few subjects, on the other hand, have risked permanent damage to their hearing in studies on the blast effects of new weaponry.

Some risks to healthy research subjects are counter-intuitive. When Stanley Milgram studied subjects' willingness to shock others when an authority told them to do so, he believed that most subjects had benefitted, because after they were debriefed they believed they benefitted from learning that they could be destructively obedient. Yet, others have asserted, notwithstanding these subjects' self-statements, that this inflicted insight was harmful.

Studies also may harm groups. For instance, when servicepersons with sickle-cell trait are studied in simulated high-altitude chambers, as a result of the findings, all servicepersons with sickle-cell trait could lose the opportunity to serve as pilots. When researchers themselves volunteer to be subjects in their own studies, they seem ideal because, more than anyone, they understand the risks, but they benefit because they add credibility to their study's purported safety.

Findings from healthy subjects may not apply to sick patients, and studies on sick patients may be misleading because sick persons may have damaged organs. The main reason healthy patients are chosen over sick patients, however, is because sick persons are more vulnerable. Hans Jonas calls this the principle of descending order of permissibility. This holds that whenever possible, persons less vulnerable should undergo research risks. When patients are adults, this might be healthy researchers; when patients are children, the healthy children of researchers.

A paradigmatic case often highlights ethical values most in conflict, though any one case, like a metaphor, both illuminates and distorts. A paradigmatic case which highlights the values likely to be in conflict when a healthy subject is used is that of a healthy, sixty-five year old man, who wishes to sacrifice the motor branch of the median nerve of his non-dominant hand as a control in the same study on his wife, who has amyotrophic lateral sclerosis. The harm in this case is small, but as Dr. Robert Levine points out in responding to this case, without both thumbs, it is not so easy to tie one's shoes or open a pistachio nut. In analyzing the case, Dr. Levine enunciates several conditions which should be met in all studies using humans, such as the lack of a less harmful alternative and potential importance of the results. Dr. Levine then asserts, however, that even if these conditions are met, permitting this research could open the door to greater harms in subsequent research. Therefore, he would not permit the study unless some authoritative national body approved it. Both these approaches, stipulating conditions which must be met and requiring procedural safeguards, are used in difficult ethical cases.

Philosopher Doug Lackey responded to this same case differently. He asserts that persons tend to believe risks are greater when the harm is small and its likelihood great as it is in this case than when the harm is large, but its probability small. Lackey asserts that what is important, however, is the product of the harm's magnitude and its probability. Do you agree? Lackey also asserts that reviewers tend to suspect normal subjects' rationality if subjects participate in research for some reason other than their own self-interest. He believes that in this case the husband's emotional involvement, as opposed to raising doubts, should be grounds for respecting his autonomy to a greater extent because the husband's desire is rooted in his feelings. This notion is heuristic. If extrapolated to research using sick subjects, it suggests that a sliding scale be used such that the more a subject's feelings were involved, the more discretion they should have in choosing whether to enter a study. This, of course, would be the opposite of Jonas' descending order of permissibility.

The view one takes depends on which analogue applies. Persons have the freedom to risk their life in race-car driving, and if they risk their life in the military, they are heroes. On the other hand, persons are not permitted to sell a kidney; if this is the closest analogue, though some persons might benefit from participating in the research, they should not be permitted to do so.

Sick research subjects differ from healthy ones in several important respects. They, unlike healthy patients, primarily want to get well. When entering phase I studies in which the safety of new drugs is determined, therefore, if their chance of benefitting is small, it is particularly important that their physicians fully discuss with them the option of their receiving palliative care or going to a hospice. I will say more about this later.

A protocol can benefit sick patients, of course, by ameliorating their illness and by offering them hope. Dr. Ruth Macklin points out, however, that the latter gain may be offset by creating false hope or, as Dr. Eric Meslin states, disappointment if the experimental treatment fails.

For many patients, the primary gain is the opportunity to be altruistic. I recall in this vein a pregnant woman with an anencephalic fetus. She wanted to gain meaning from her loss by finding a way in which her fetus's organs could benefit another. I also recall a physician who had aplastic anemia. He chose to die rather than to continue receiving blood and platelet transfusions indefinitely. He found some solace from donating his body to researchers whom he hoped, as a result, could understand his condition better and perhaps use this knowledge to benefit others.

The use of a randomized clinical trial is the ideal way to determine whether a new treatment is most effective. Requiring patients to serve as controls, under some circumstances, however, is problematic.

The use of a placebo control when patients have high blood pressure, for instance, can create more than minimal risk. If anti-hypertensive medications are stopped, patients may have a stroke or die. This risk can be limited by allowing only patients with milder hypertension to participate or it may be possible to maintain patients on their medications, and add on the new drug or placebo. This add-on approach renders the research less sensitive, but it also has some advantages. It may, for instance, make it less likely that positive effects of the new drug would not be identified. If, for example, the patient's disease had flare-ups, the benefit of a new drug might not be apparent.

The harm created by using a placebo or best standard treatment may be more problematic when the illness can be lethal. An often cited study compared adenine arabinoside with a placebo in the treatment of herpes simplex viral encephalitis. The percentage of patients dying without the new drug was seventy percent, and thus there was considerable reason to believe that adenine arabinoside would be efficacious. Some argued that this study should not have been done.

This problem is greater when an illness is uniformly fatal. In this situation, it is difficult to use a placebo or ineffective best treatment when a new drug appears promising; researchers may feel great pressure to use the new drug even though it has not been adequately studied in animals. For example, when Pasteur was developing a treatment for rabies, Pasteur resisted a plea to treat a bitten child on the grounds that his method had not been securely established; just eight months later, he treated a patient but experienced, he wrote, "profound anxiety." This same problem occurs today. In a March 1989 issue of JAMA, four cases were reported in which partial surgical evisceration, replacement with organs from cadavers and immunosuppressive chemotherapy were carried out on four desperately ill children. Dr. Francis Moore commented wryly, "I believe that this procedure should not be performed again until it has been shown that there is a palpable likelihood of success."

In cases such as this, one approach which might be considered is a "play the winner" strategy. In this approach, treatments are randomized, but the chance of a patient's receiving an experimental or standard treatment or placebo varies depending on the outcome of each previous patient. This approach is particularly useful when the outcome of the intervention, life or death, will rapidly be known, and when one intervention appears likely to be much more effective than the other. Alternatively, historical controls, literature controls, or observational controls can be used. This last approach may be most effective when physicians already are using a new drug or treatment which has not been scientifically validated.

An exception to the arguments against using a placebo or ineffective best treatment as a control when patients' illness is uniformly fatal and a new drug appears promising occurs when only small amounts of a new drug are available. This occurred, for example, when streptomycin was first available to treat tuberculosis(TB). Two hundred veterans with TB were moved to Washington and half were given the then-best therapy and half were given streptomycin. After the streptomycin proved a clear winner, it was given to the survivors who had served as a control. But during the last six months of the study, streptomycin was more readily available. Dr. Meier, the chief researcher stated, "I found myself troubled by the ethics of continuing to withhold streptomycin from the control subjects."

This same dilemma recurs today. If only small amounts of a drug are available, they should be used in a randomized clinical study and should be used "compassionately" only if an adequate, randomized clinical study is full or there is excess drug for patients who do not meet inclusion criteria. Also, if additional studies are necessary and sufficient drugs are available, these studies should be conducted concurrently.

What if there are ample amounts of the new drug but a randomized clinical trial has not yet been carried out or completed? Surely, most patients would seek the new drug from a private physician rather than enter a study and take the chance that they would serve as a control. Benjamin Freedman has suggested that one response to this situation when it involves patients with AIDS is to establish AIDS Clinical Trial Units at which experts could determine on a local level what randomized clinical trials should be performed and who, if anyone, outside the study should receive the promising drugs.

The core ethical questions underlying these situations are: when is a randomized clinical trial necessary?; and when, in the absence of a trial, should persons with fatal illnesses be precluded from having access to a promising new drug? In some cases, a trial may be impermissible because the potential harm to subjects in the control group is too great. A study in Britain was proposed, for example, in which some women who had previously given birth to infants with neural tube defects would receive vitamins, hopefully to prevent this outcome from reoccurring, but those in the control group would not. This study was not permitted.

In other instances, it is ethically permissible to begin a study but not to continue it because the preliminary findings suggest that one treatment is far preferable to the standard treatment or placebo. The National Heart Lung and Blood Institute, for example, curtailed a multi-center study comparing propranolol with placebo in patients who had an acute myocardial infarction because the patients receiving propranolol had showed a 26 percent lower mortality rate. This difference was consistent both across clinical populations and within patient subgroups.

In some studies, investigators moderate their own data, but in double-bind and large studies, data monitoring committees usually assess the data periodically to determine whether the research should be stopped. Data should not be monitored too frequently, however, because if it is the probability will increase that one treatment group would be significantly different due to chance alone. The criteria for stopping a study also, ideally, should be established before the study begins in order to avoid suspicion that the outcome has been manipulated.

When should a randomized clinical trial be initiated? Chalmers contends that preliminary data obtained in an uncontrolled pilot study inevitably tips the balance in favor of one intervention over another; therefore, a randomized trial on a new drug should be carried out, if possible, with the first subject. Chalmers' experience prior to taking this position is instructive and moving. He states, "I treated many patients incorrectly because I did not know about controlled clinical trials, and that is why I decided to devote my life to convincing students and physicians of the crucial importance of randomized clinical trials." Schaffer similarly believes that there is an ethical conflict between giving patients' interests first priority and enrolling patients in a randomized clinical trial. He believes, however, that so long as patients are adequately informed, randomized clinical trials are justified, in part because they further the social welfare.

Other writers have provided additional justifications of a social-welfare centered ethic. Miller states that this priority might be justified because patients should reciprocate for the sacrifices past patients have made which benefitted them. He then points out, however, several problems with this claim. Patients now living, for example, did not ask patients to make these sacrifices. Gifford argues that knowledgeable persons would freely agree to a social-centered ethic. He states that a contract model, however based, is the "kind of justification that is needed, has some plausibility, and holds the [most] potential of being able to be fleshed out into policy...."

The late Benjamin Freedman and Dr. Levine argue that patients' best interests are not sacrificed when they enter a randomized clinical trial because there is no scientific proof that one treatment is better than another. They contend that while theoretical equipoise is an exact balance between the two arms of a therapeutic trial, clinical equipoise represents genuine uncertainty among medical experts.

For this same reason, subjects' interests are not compromised if they are not informed of emerging trends after they enter a trial unless some subject underwent a serious side-effect which has not been anticipated. Subjects should be informed of emerging trends before they enter a trial, however, to fully respect their autonomy.

Whatever position one takes, some persons clearly place value on treatments which have not been scientifically validated. Parker and Sachs, for example, presented a paper at the AIDS Conference in Montreal, in which they argued that foster children who were infected should have the same access to therapeutic trials as other children.

What is the appropriate analogue? Quarantine, in which individual patients' interests are sacrificed for the good of others? Or is a randomized clinical trial not a sacrifice, as Dr. Freedman and Dr. Levine argue? Or is it both?

In the military, several aspects of doing research merit recognition. First, harms to which military soldiers could be asked to expose themselves may be exceptional since research may involve dangerous weapons. Subjects may be needed, for example, to study the effects of new weaponry on persons using it. If subjects fire new weaponry, blast overpressure might cause damage to their hearing which could not be determined from animal studies alone.

The rationale of military necessity could also make it compelling for subjects to expose themselves to higher risks. If a potentially hostile nation is conducting research on new weapon, it would be necessary for this country to do likewise. Because soldiers are willing to make sacrifices for their country during combat, researchers consider them willing to risk greater morbidity as subjects of experimentation. This conclusion, even if true, would be ethically improper. Soldiers already make exceptional sacrifices for society and societal burdens should be distributed equitably. For this reason, conscientious objectors during World War II served as subjects in starvation experiments and some have suggested that prisoners, who do not risk being killed in battle, should be drafted during war as research subjects. Do you agree?

A different kind of risk regarding soldiers is the possibility that researchers asking them to serve as experimental subjects will exploit their inherently coercive situation as soldiers which may be due to the nature of their military duties. Soldiers could volunteer because this will be noted in their personnel file and might they believe that this would help them to be promoted to a higher rank. Finally, as mentioned already, some soldiers could volunteer because they have neurotic needs to make sacrifices for others, a need which could be expressed when soldiers repeatedly volunteer for hazardous missions during combat. In each of these situations, conditions exist in which the soldier's external environment or internal psychological state might interfere with the soldier's being able to make a truly free choice to volunteer.

At the present time, unless a national emergency occurs, military experimentation must conform to all standards required in civilian federal institutions. Experimentation in military settings is subject to the strictest scrutiny and, therefore, these issues may be of more theoretical importance than practical significance.

Are there experiments in which military physicians should not participate? Why might military physicians refuse to participate in certain types of research? Generally, some fear that certain kinds of research could be used to develop weaponry which would cause excessive suffering to soldiers or harm large numbers of innocent civilians. Weapons considered in these categories are mostly those which produce fire, chemical, and biological. Compare these considerations regarding military research with one which occurred a few years ago in regard to whether anyone should conduct research on recombinant DNA.

There is a distinction between research that no one should do because of its potential harm to others and research that physicians should not do because in becoming physicians they have vowed to do no harm. Theoretically, the latter argument presupposes that physicians have obligations as members of the medical profession which go beyond those of other citizens who are not physicians.

Presumably, physicians' extraordinary obligations stem in part from medical codes such as the Hippocratic Oath in which physicians promise to do no harm and to preserve life. There is some question, however, concerning the degree to which this and other oaths might have been based on more expedient concerns such as attracting patients to maintain physicians' livelihood.

This claim, that physicians should not participate in research that could harm others alternatively may be based on the independent moral concern that doctors have made an implicit promise to further the general public's as well as individual patients' health. Some doctors argue that physicians have a duty to practice preventive medicine and that this promise obliges them to speak out in favor of nuclear disarmament because this is the only means by which patients morbidity and mortality from nuclear war can be prevented.

This notion of physicians' having supererogatory duties beyond those of an ordinary citizen is far-reaching. Some believe that physicians should risk exposing themselves to AIDS when treating these patients in ways that non-physicians might not, and conversely, that physicians should not give lethal injections to criminals condemned to death which non-physicians might.

When a person takes an oath both as a physician and as a military officer, does a conflict exist? If it does exist, is the latter oath a disclaimer of the military doctor's medical obligations as a physician? Are military physicians free from supererogatory obligations they might have on the basis of being a physician? Does the military physician have obligations to society to do research on biological weaponry even though this weaponry could harm "innocent" civilians because this research is necessary to protect this country from hostile nations taking it over? Should these obligations to society override a military physician's obligations as a physician?

At what point is this argument no longer valid? The point might come at which the means of war used by a country might be such that one would no longer want to be a citizen of that country. Then, however, should the military physician refuse as a citizen or as a physician? Some would argue that one should never refuse as a citizen because a country is justified in using any means of war so long as the ends are morally just.

Two new issues involving research and the military arose during the war in the Persian Gulf: Should the military give servicepersons compounds which appear highly beneficial against biological or chemical warfare, but which have not been fully tested? Should servicepersons have the right to consent or withhold consent to this treatment?

All military physicians should be familiar with the ethical arguments underpinning the military's position on these issues. Thus, the article "Treating the Troops" is required reading for this session. I have not included an additional, dissenting article on this issue, but if you are interested, this and other readings on this subject are on reserve.

You may be wondering how it is that compounds that would protect servicepersons from biological weaponry have not been fully tested. The lack of ethical justification for testing defensive agents against biological and chemical weaponry using human subjects should be self-evident.

Consider these related questions: What other agents which have not been fully tested should also be used during combat? What if there is an agent which appears highly effective in preventing death due to endemic disease or serious injury? Should it be made available to servicepersons in combat even when it could not be made available for civilians? How would you argue for or against such an agent's use?

Returning to agents considered in the Gulf, what obligation do you believe the U.S. had to make these agents available to servicepersons from other nations and/or civilians in Kuwait? What about captured enemy servicepersons? What might the Geneva Convention suggest in regard to these questions? Do you agree?

A recent most controversial issue is whether the US should apply its own standards when doing research in other countries. To do so furthers equity but it also has been described as "cultural imperialism" when another country would want research conducted but the US will not allow it. This question arose in regard to doing research with AZT to prevent pregnant women from transmitting HIV to their infants in Africa. The amount of AZT used would have been much less than the amount found in the US to be somewhat effective. Some subjects would have been taking a placebo. Yet, if the small amounts of AZT proved effective, it is possible poorer nations in Africa and elsewhere could afford these much smaller amounts of AZT to help prevent transmission of HIV from pregnant woman to their offspring. Do you think the research should have been done? Some readings about this issue are placed on reserve.

The smaller amounts of AZT, in fact, were showed to have preventative effects in Thailand.

Animals' interests are being recognized to an increasing extent. Regardless of where you stand on these issues, there are some ethical arguments and their criticisms with which you should be familiar.

A first issue involves the question of speciesism. According to one argument, there are no morally relevant differences between higher animals and human beings which would warrant humans subjecting animals to research which they would not also conduct on human beings. If no morally relevant differences exist which would justify a discrepancy, and research is done on the basis of animals' being a different species alone would this be as repugnant as racism or sexism?

Those making this argument sometimes also use this logic: Is not a dog or cat more intelligent than an anencephalic newborn infant, a severely retarded person or a comatose patient? Since it is, intelligence therefore cannot be a morally relevant difference to justify the use of dogs and cats in research. Or, if dogs and cats can be used, so should anencephalic newborns, severely retarded humans and comatose adults.

Consider aspects of this latter assertion. Human beings, it could be argued, should not experiment on comatose persons, for several reasons. First, humans place more value on a comatose person than an animal because comatose persons are still living humans and at a previous time in their life experienced significant relationships with others. Not to do experiments on humans under these conditions respects their dignity. In addition, since any person may become ill or injured and then enter into a comatose state, if experiments are done on comatose individuals, some humans who anticipate this possibility might be fearful.

What more can be said? Reconsider the concept of intelligence. A dog or cat may have more intelligence than humans in certain circumstances, but humans as a species have capacities which animals do not: capacities to create culture and transmit culture from one generation to the next, to use language, to reflect upon oneself, to anticipate death and to form moral priorities. Or do they? Whether these differences warrant treating humans differently from animals is open to debate. Some argue that these very differences require humans to protect animals.

A different argument involves animals having rights. There are many arguments concerning how one can come to have a right, but the one most commonly cited is that one voluntarily chooses to give up some freedoms to gain certain rights in return. One gives up the right to assault someone else (that is, accepts criminal sanctions being taken) in order to have the right to be free from being assaulted by another.

If one uses this theory of rights, an individual claiming to have a right must have the capacity to make such a contract. Thus, using this sense of rights, it makes no sense to talk of animals having rights, because animals lack the capacity to give something up in order to get something in return. Animals may engage in behaviors which suggest they have this capacity. They may not eat off a human's plate when a human is eating to gain that human's favor and affection, but this restraint is plausibly understood by a behavioral paradigm such as conditioned fear or learned avoidance and reinforcement.

An argument often raised against higher animals, such as apes, being excluded from research is a slippery slope argument. According to this argument, if humans must respect the interests of higher animals, then they must also respect lower animals, insects, and even plants. This argument is a poor one. There may be no fully satisfactory way to make an ethical distinction between some animals' interests and others. Nonetheless, this hardly means that no relevant moral differences exist between doing research on apes and insects or plants.

Compare the question when, if ever, abortion should be permitted. If abortion is ethically permissible at any time, there may be no satisfactory arguments for establishing a time in fetal development before which abortion should be permitted and after which it should not. The line is presently drawn legally at viability, but various other points could be used. Thus, the time at which abortion should be allowed--if it should be allowed at all--may be arbitrary. Yet it does not follow that because this line is arbitrary, no line should be drawn.

It is obvious that using animals as research subjects can benefit both humans and animals, but that using animals as subjects of experiments causes animals' suffering. There is a more subtle argument for limiting animal research worthy of reflection. This argument claims that humans who conduct research which causes animal suffering may become emotionally less sensitive as would persons who perform or know of human atrocities and fail to act to prevent them. Is this argument valid? If so, how much weight do you think this moral concern should have?

A final question of a different sort is whether the military shouldd use animals. An example involving dolphins is the third required reading. You will discuss this question in your discussion groups.

FACILITATOR'S GUIDELINES FOR DISCUSSING CASES

SESSION IV - SEPTEMBER 15,1998

Case 1 - The major issue in this case is whether or not it is ethically appropriate to employ placebo controls when there is some evidence already available which suggests that a patient will respond better to one treatment than another. One view might be that the effectiveness of the surfactant could be established by providing it to all RDS patients such that conscientious record keeping could then provide data which would reasonably establish its efficacy. The alternate position is that regardless of prior studies, there is no data concerning the efficacy of surfactant in this situation to support the clinical use of the drug and that furthermore there may be unanticipated serious adverse effects of surfactant such that this drug's efficacy or toxicity should be established by performing randomized studies first.

This case, as the next one, emphasizes the potential conflict inherent when a physician engages in medical research. As clinicians their primary responsibility is to the individual patient. The physician implicitly promises to do whatever he believes is in the best interest of the patient. According to one point of view, if a clinician were treating a patient with an illness for which no treatment had proven effective, and if the clinician knew of just one case in which a patient had responded to an experimental treatment, the clinician would recommend that proposed treatment.

The other view holds that before data has reached levels of scientific certainty, there is no proof that it is efficacious and offering it to a patient as if there is such proof is misleading. This assessment is to some extent a value judgment in itself, and one with which patients may differ.

The facilitator might help the students appreciate these different theoretical views by changing the figures hypothetically regarding the improvement of premature infants taking surfactant. Suppose, for example, that there were ten neonates with RDS taking surfactant and they had showed a ten percent mortality rate and no morbidity; one percent? The students should recognize that as the figures change, the argument for conducting the randomized study becomes harder and harder to maintain.

This particular case is complicated also by the fact that the research subject is an infant and standard treatment has substantial mortality and morbidity. Questions the students might consider are: who should consent on behalf of the infant? When the infant is premature, can the parents make an objective judgment? If the parents differ in their views, could their differing adversely affect their future relationship? Since the infants cannot themselves consent, how should the researcher err, i.e., should they stop the experiment earlier when there is some evidence of "substantial improvement"? What role, if any, might an ethics committee play in deciding these issues? Would an ethics committee have more objectivity than the parents and physician and/or would it interfere with the physician and parents making the decision together?

Case 2 - This case poses some issues raised by the previous case, but is distinguishable in two regards. First, in AIDS the fatality is a hundred as opposed to twenty percent. The students should be asked the degree if any to which they believe the patient's expected fatality should affect the degree to which a study is carried out, short of scientific certainly being reached. The second difference is that in this case the patient is an adult. Thus the students can be asked the extent to which adults should be given information concerning the results of this study, as they are acquired. Giving patients ongoing access to data from the study might reduce its effectiveness. At what point would it be ethically unacceptable to withhold data from the patient for purpose of this study?

Case 3 - Raises on a policy level some of the issues addressed in Case 2, but it also raises a most important question regarding the equity with which patients "in need" of experimental therapies can have access to them. A fact sheet on the FDA's final rule on the treatment use of and charging for investigational drugs is appended to this guideline for facilitators' interest.

Limiting access for an experimental drug may violate justice, and limiting access by allowing the sponsor to charge will restrict the benefits and burdens of research to wealthier persons. The students might consider this question: If public monies have contributed in a substantial way to the development of the drug, as, for example, by funding basic laboratory research which enabled the chemical nature of the drug to be formulated, would this public "investment" strengthen the argument that all should have equal access to the drug?

The students might be asked in this regard whether patients with AIDS who had a strong history of drug abuse should have had initial access to AZT. The authors of the first required reading expressed themselves strongly on this issue. Some believe that even if drug abusers are excluded according to a "medical" rationale - namely, their being "unreliable" and "noncompliance" - excluding them is nonetheless unjust.

Finally, the students might be asked how they believe a policy allowing sponsors to charge for experimental treatment would affect sponsors' approach to research. Would these sponsors market new therapies more rapidly? Would formal clinical trials become more difficult thereby delaying the availability of proven treatments to the public?

Case 4 - This case in intended to give students the opportunity to consider primarily two questions: is there any research potentially useful to the military they should not do, and, more generally, should they ever compromise their personal ethical beliefs for the sake of their spouse or children, the military or society.

A first issue which might be discussed is whether or not there are some kinds of research that shouldn't be carried out by the military. There has been consensus in our history that certain weapons such as the dum-dum bullet should not be used in warfare. Students might be asked what other kinds of weaponry, such as weaponry that can cause excessive pain or biological organisms whose destructive effect cannot be limited, should also be forbidden.

An issue likely to be raised in this discussion is whether or not any weapon, no matter how heinous, is justified if necessary to protect this country. If weaponry would be heinous if used offensively, would a physician be justified in doing research on it for defensive purposes?

Case 5 - Before considering this one case regarding the use of animals in research, students might first be asked to share emotional responses they have already experienced when doing procedures on animals. The facilitator might then raise the general question whether humans can justifiably impose suffering on animals when they use them as subjects for research. Some students may raise the argument that since animals are killed for food, it is a "lesser evil" to use them as subjects for research. The students could be asked if they see any distinctions between using animals for preparing food and as subjects of an experiment. If some students believe that subjecting animals to suffering is ethically unjustifiable, you might ask them whether politically it might be more possible to curtail one societal practice than the other, and if so, whether this question has ethical relevance.

The case itself raises first the factual question whether "higher animals" such as a pig can experience psychological suffering. The facilitator might ask the students whether they feel that there is uncertainty regarding this question and if so, who should have the "burden of proof", the person claiming that the animal suffers, or the person claiming that it does not. One approach to inferring animal suffering involves extrapolating from human experience. The Federal Interagency Research Animal Committee has recommended that in caring for animals an investigator should consider that any procedure that causes pain or distress in human beings may cause pain or distress in other animals.

The question may then be asked, even if the pig would suffer, are there some kinds of experiments so important to humans that the infliction of even substantial suffering would be justified? If, for example, an experiment on a chimpanzee - or pig, if pigs were a more appropriate model - could result in a cure for AIDS, would the chimpanzee's experiencing considerable suffering be justified? Some have claimed that such experiments should be performed on anencephalic newborns or permanently comatose adults. The students might be asked what they think about these proposals. The chimpanzee is more "intelligent" and can "relate" more than the other two. Yet, an anencephalic newborn is human and can evoke "bonding" feelings in its parents and if comatose adults were used, humans could fear that if they became comatose, they could become research subjects. This last concern would be overcome if humans could be used as research subjects only if they gave prior consent when they were competent. The students might be asked if they would be willing to give such consent, and if not how, if at all, they would justify using a chimpanzee.

Case 6: This case raises the ethical issue of whistle blowing, in general, and whistle blowing by a USUHS student, in particular.

Initial considerations are the degree of harm to the animal and the certainty that this harm occurs. The greater the harm and its certainty, in general, the stronger the obligation to take some sort of action.

A further consideration in this instance, is the degree, if any, to which the rabbits' care deviates from the standards which are required. Here there is again a question of potential whistle blower's certainty that there is a deviation from the standard. The head of the lab may know better than the student whether there is a deviation.

If the laboratory conditions are within the present standards, to the degree that these standards have been lawfully enacted, they represent standards accepted by the community. In this instance, students' ethical grounds for objecting would be the same grounds for objecting to the standards themselves.

If there are deviations from the standards, the students might believe that they should object in all instances or alternatively only when they think the deviation is extreme. In the latter instance, the student's idiosyncratic value bias may play a role. This raises an important question the student should discuss. Ask them to assume that two students observe the same deviation, but both have highly different values regarding the importance of animals' interests. Should they pursue the identical action or is their value difference a valid basis for their pursuing different actions. That is, should protest be based on objective or subjective criteria?

The above considerations are generic. The question for USUHS students, in particular, is whether they have any additional obligation not to blow the whistle as a result of their identity as a USUHS student. Have they, for instance, made some implicit promise in coming here to USUHS or the military which would give them a stronger obligation to remain silent? Related concerns are the potential harm to USUHS and/or the military and its probability. Further questions the students might discuss are the potential harm to themselves if they blow the whistle and its probability. At what point, if any, would it be praiseworthy and at what point, if any, ethically obligatory for them to act regardless of personal sacrifice?

The facilitator should explain before leaving this case that there are innumerable actions students could and should take "within the system" to resolve problems such as this should they arise. Thus, it is extremely unlikely that the question of whistle blowing would ever arise. Still, when working within the system brings no results, the whistle blowing question arises. It is therefore important to discuss.

Case 7 - Whether dolphins or other animals should be used as weapons is likely to depend on how students regard animals.

If they regard animals as fundamentally vulnerable - like children or the mentally ill - and thus - warranting society's protection - they may believe that their use in combat should be precluded. Students could argue plausibly that animals should be used as food and research subjects but not in combat on grounds of necessity.

This argument would tend to presume that the use of animals is not critical to the combat effort. This is, of course, a factual question which students may find it productive to debate.

A different, additionally relevant consideration would presuppose that the use of animals is justifiable in some cases. Then a first important concern would be the degree, if any, to which these animals would suffer. Again this is a factual question. The second question of concern would be the degree to which animal use would further the combat effort in a way in which there were no comparable alternatives. This also is a factual question. Assuming that there is agreement on both questions, the analysis would then be analogous in many respects to that of the degree to which humans should be used as research subjects.

Students may, on the other hand, argue plausibly that the use of animals should be permitted in combat, even though not permitted in research. The ground would be again that of necessity. Namely, the use of animals may be necessary to win a war and the outcome catastrophic if the war is lost.

Session IV - Research

September 15, 1998

Case 1

You are a member of an Institutional Review Board evaluating a proposed research project of the treatment of respiratory distress syndrome (RDS) in premature infants. The principal investigator has explained that RDS is associated with a mortality rate of greater than 20% in neonates treated with standard ventilator therapy, and that 50% of the survivors have serious pulmonary and/or CNS sequelae. The proposed trial will employ the prophylactic use of surfactant isolated from amniotic fluid in an attempt to prevent or mitigate RDS. Preliminary uncontrolled studies of this material in a small number of patients have demonstrated a "substantial improvement" in survival. The investigator proposes randomly assigning infants to the control (placebo) or treatment groups.

1. Although the treatment is not of proven efficacy, there is evidence to suggest that it is beneficial; in that circumstance, is it ethical to conduct a randomized placebo-controlled trial in such a serious disorder or should all RDS patients be offered this treatment?
2. A fellow panel member proposed that if a placebo-controlled trial is not performed, clear evidence of efficacy might not be obtained and the availability of a potentially beneficial treatment will consequently be delayed, ultimately resulting in a greater number of deaths. That is, the benefit to society must take precedence over potential benefit to the individual. Do you agree?
3. Is the issue of informed consent in placebo-controlled trials different in any way for neonates and infants than for older children or adults

(Adapted from a case submitted by Dr. Thomas V. Holohan)

Case 2

Zidovudine (AZT) was recently approved for use in the treatment of certain complications of AIDS. The initial clinical trials compared a treatment group with a placebo-controlled group. The trial was stopped before all patients had completed the planned 24 weeks of therapy when statistical analyses of the trial indicated a significantly higher death rate in the placebo group (19/75) than in the AZT group (1/85).

1. Is it ethical to employ a placebo group in treating a uniformly fatal disease for which there is no effective alternative therapy? Is this issue in any way different from that in the previous case?
2. Did the conduct of the AZT trial (i.e., stopping the controlled trial as soon as there was statistical evidence of effectiveness) adequately address this issue?

Case 3

The Food and Drug Administration (FDA) has adopted a rule permitting the use of investigational drugs for treating patients outside of a clinical trial in certain circumstances. That is, in serious or life threatening diseases, the sponsor of a drug undergoing clinical trials for approval for marketing may request the FDA to make the drug available to practitioners for treatment of their patients outside of the clinical trial setting. The manufacturer may also request permission to charge for the drug, the charge being related to recovery of costs for manufacture, R&D, and handling.

1. Is appropriate to allow unproven drugs to be used as treatment?
2. Is the overall effect of the rule beneficial on the basis of its providing of wider access to potentially useful treatments?
3. Should an unproven drug be made available in a setting wherein there is little likelihood of obtaining reliable data as to its effectiveness?
4. Is it appropriate to allow sponsors or drug manufacturers to charge for an experimental treatment? Is it unreasonable to expect that sponsors sustain all the costs incident to obtaining marketing approval?
5. Is charging for an experimental therapy just? Is it ethically imperative that access to experimental treatment be available on an impartial basis to all?

(Adapted from a case submitted by Dr. Thomas V. Holohan)

Case 4

Having recently completed his Ph.D. degree in chemistry, George has not been able to find a job. His family has suffered from his failure since they are short of money, his wife has had to take a full-time job, and the small children have been subjected to considerable strain, uncertainty, and instability. An established chemist can get George a position in a laboratory that pursues research in chemical and biological warfare. Despite his perilous financial and familial circumstances, George feels that he cannot accept this position because of his conscientious opposition to chemical and biological warfare. The older chemist notes that while he is not enthusiastic about this project, the research will continue whatever George decides. Furthermore if George does not take the position it will be offered to another young man who would probably pursue the research with alacrity and promptitude. Indeed, the older chemist confides, his concern about this other candidate's nationalistic fervor and uncritical zeal for research in chemical and biological warfare in part led him to recommend George. George's wife is puzzled and hurt by George's reaction since she sees nothing wrong with such research. She is mainly concerned about the stability of their family and their children's problems.

(Taken from Tom L. Beauchamp and James F. Childress, Principles of Biomedical Ethics. Adapted from Bernard Williams, "A Critique of Utilitarianism," in J.J.C. Smart and Bernard Williams, Utilitarianism For and Against, Cambridge: Cambridge University Press, 1973, pp. 97-98.)

1. What do you think George should do?
2. Suppose he were a physician in the military asked to conduct this research? What do you think George should do?
3. Are there kinds of research in chemical and biological warfare that should not be carried out?
4. If potentially hostile nations were carrying out chemical and biological research, would you answer differently?

Case 5

Anesthesia in patients with povolemia presents special problems. The objective of this project was to improve the anesthetic management of moderately and severely injured patients suffering from blood loss. The animal model to be used was the paralyzed, but unanesthetized pig.

When the animals were conscious but paralyzed and ventilated, the following procedures were carried out:

1. Measurement of hemodynamic and hormonal variables before and after hemorrhage of 30 ml/kg. The animal's response to this intervention might be expected to be shock and feelings of panic.
2. Effects of the anesthetic agents enflurane, halothane, isoflurane, nitrous oxide, ketamine or thiopental on hemorrhage response. This will require manipulation of the intratracheal tube and again the animal may experience panic.
3. Paralyzed, but conscious pigs, may be stressed by these procedures. A hemorrhage of 30 ml/kg is not trivial. How can "stress" in a paralyzed animal be assessed by the experimenters? Can pain, physical and/or mental be experienced by paralyzed animals? If you feel there is no way to answer this question should you presume the existence of pain or its absence?

Should the experiment be done if conscious pigs are used?

Case 6

Assume that you observed these conditions here at USUHS - rabbits living in poorly maintained cages encrusted with dried urine, feces, and spider webs and not fed regularly. What should you do?

Assume that you have expressed your concern to the appropriate person responsible for the labs and animals' care but that these conditions have not been ameliorated. How far should you go? Do you believe that an anonymous call to a newspaper or politician would be ethically justifiable? If not, under what circumstances, if any, do you believe such a call would be justified? If there are no such circumstances regarding animal research, would there be any regarding human research? What about the notorious experiments discussed, for example, in the introduction to this session?

Case 7

Do you think dolphin's and/or other animals should be used as weapons? Why or why not? Are there any limits you would establish? If so, what are they?