psychiatry: methods and processes for investigation of drugs

PSYCHIATRY: METHODS AND PROCESSES FOR INVESTIGATION OF DRUGS

Leo Alexander, M.D.*

Tufts University Medical School Boston, Mass.

The effects of psychotropic drugs on mental or emotional disease can be studied in two ways. The first is by a relatively short-range determination of the direction in which the patient population changes as an effect of exposure to certain drugs. This short-range investigation has certain drawbacks, the chief of which is that these studies are rarely pursued to the point of actual recovery of the patient. These studies usually limit themselves to the establishment of a certain amount of scorable or otherwise definable improvement within a certain limit of time. The prime requirement is proper matching of the patients and the controls. The best way to obviate this difficulty is to let every patient be his own control by the double-blind crossover technique. The patients are exposed to a period of trial treatment, half of which, randomly selected, is spent while receiving the drug under study, and during the other half of the observation period, the patients receive either placebo or another drug of known effect to which the new drug is to be compared. The method is truly double blind, in that neither the patient nor the active investigator know what the patient is receiving at any particular time, the code being kept by a nonparticipating physician. The chief drawback in the use of this method in psychiatry, even under those optimally controlled conditions, is, of course, the limitation of the result obtainable within the specific limitation of time. The amount of improvement obtainable during such a circumscribed time, unless it should result in complete recovery, does not have sufficient relevance to the problem of recovery from mental or emotional disease, because drugs may produce improvement without necessarily leading to social or complete recovery, which is the ultimately desirable goal of all treatment in mental and emotional disease. At an extreme of nonrelevance, patient populations in certain research wards of chronic hos- pitals may be treated and allowed to improve with one set of drugs after another without ever coming to the point of rehabilitation and discharge.

A method of greater relevance to the practical problems in psychiatry is to use drugs to bring about social and/or complete recovery of a patient without narrow or circumscribed time limits. This is often not obtainable with any one set of drugs except in a relatively small number of cases (about one-third); hence the investigator using this method has to be prepared to give one drug after another until the patient actually recovers. The controls in this kind of study are provided by the large number of the sample that is necessary. The double-blind crossover can give some rough ideas of certain target effects of the drug within one or two months’ time and in a relatively small number of patients, such as 20 or 40, while the other method of study is structured along the actual way in which therapy progresses, but requires several hundred patients and

*Formerly, Consultant to the Secretary of War and the United States Chief of Counsel for War Crimes.

344

Alexander: Psychiatry-Investigation of Drugs 345

controls, thoroughly analyzed according to each phase of treatment. In other words, in such a study the results obtained in the course of the first treatment phase cannot necessarily be compared with treatment phase 2, because the patients participating in phase 2 are already a selected group, based on the fact that they failed to respond in drug treatment phase 1.

In carrying out such a study, my collaborators and I have found rather significant reflections of previously ineffective forms of treatment. In a study of 767 consecutive cases so studied, we found that in a group of unselected cases of depression (735 cases), the drug treatment administered during the first treatment phase produced an average recovery rate of 26%, which varied remarkably little among the various types of antidepressants used. It was 25% for tricyclics, 30% for monoamine oxidase inhibitors, and 27% for meprobamate in combination with benactyzine.

Electroshock treatment administered as the first treatment phase produced a recovery rate of 60 percent. However, when electroshock treatment was administered after failure of treatment with tricyclics, only 36% achieved recovery with electroshock. If electroshock was administered after failure to respond to meprobamate combined with benactyzine, 61 % of the patients achieved recovery with electroshock. This is obviously due to the fact that the biochemical and psychophysiological mechanisms of recovery with tricyclics and electroshock are somewhat similar: hence patients who failed to recover with tricyclic drugs may be presumed to have also a lessened capacity to respond to electroshock. However, after failure to achieve recovery with meprobamate and benactyzine in combination (the target symptom of which is different from that of the tricyclics and of electroshock in that it largely affects ruminative thinking and behavior), the patients respond to electroshock at the same rate as untreated patients, because they obviously have not been screened; such treatment has not resulted in the selection of electroshock-resistant cases. It is therefore very important in the evaluation of any treatment result to study previous modes of treatment, because such previous treatment may exert a selective effect upon the patient population that will influence subsequent treatment. Unfortunately, relatively few drug studies reported in the form of the usual short-range studies of the pilot type have made allowance for this particular contingency.

As to the ethics of this study, it is certainly my conviction that the method is well adapted to the application of all the safeguards of the protective codes of ethics that have developed since the formulation of the Nuremberg Code.' As the author of the memorandum to the war crimes court 2, from which the Nuremberg Code was derived,' I can vouch for the fact that I have taken all these aspects of experimental studies encountered during many years of experimental research with human beings into account, and I have never subsequently had the slightest difficulty in conforming to all the ethical requirements of the code. It is often assumed that it would be difficult to explain the double-blind crossover and the use of placebos to patients participating in such studies. I have not found this to be so if the method is adequately explained and interpreted to the patient. The double-blind crossover study can be candidly explained without interfering with the unbiased attitude of the patient. We tell the patient that over a trial period of, for example, two months, he will receive the drug to be tested in varying degrees of strength, and that one set of pills he will receive will contain none of the drug at all. He is told that if he were to be given the drug straight in the open, it would take a year or more to find out whether it was helpful, and he and others would have to take it for all this time,

346

possibly unnecessarily. But in this manner, neither he nor we knowing when he took the real drug, we could settle this question fairly promptly. If he should actually do better on the pills containing no drug, we would know that his illness was on its natural way to recovery. If he should experience any discomfort not promptly relieved by discontinuance of the pills, we could always break the code and take care of him on the basis of this information.

We found that this explanation was not only helpful in establishing a collaborative and mutually understanding attitude, but actually enhanced the patients’ interest in the outcome and sharpened their self-observation. Many patients would take particular pride in figuring out when they took the real drug and when they took the placebo. Needless to say, there was no consistent ability on the patients’ part to differentiate the drugs from placebo, since side effects were usually not sufficiently consistent to give the nature of the drug away. One drug that we studied in this manner happened to have been Parnate. Enlightened consent is therefore very feasible, even in the case of study of placebo controls, and the use of placebos is certainly no obstacle to the ethical rule of enlightened consent. All other aspects of such drug studies conform to the general aspects of ethics of experimentation on human beings. In order to outline this in more full detail, I should like to review briefly here the origin of the Nuremberg Code as well as its elaboration by the justices of the war crimes court.

Because the nature of the criminal experiments which formed the subject matter of the Nuremberg court is sufficiently well known, it does not need to be reviewed here. A recent review was published by the present author in 1966.4 It is often assumed that the war crimes court concerned itself only with non- therapeutic experiments. This is certainly not the case, since the exposure to cold experiments, for instance, had a therapeutic aspect, as did the typhus experiments and others. The only point of difference is the fact that the pre- dicament to be solved therapeutically had been produced as part of the experiment and was not a naturally occurring disease. But it is my opinion that the Nuremberg Code covers all these contingencies in a much more specific manner than subsequent formulations such as the Helsinki resolution. I am therefore fully in accord with our present regulation by the National Institutes of Health, which is based essentially on the Nuremberg Code rather than on subsequent variants.

The definition of ethical experiments in human beings was necessitated by the fact that counsel for the defendants accused of the crimes under investigation undertook the only logical line of defense; namely, to make an attempt to obscure the issue by presenting data on medical experiments in other countries in order to prove that any medical experimentation on human beings was in itself a criminal act or an act likely to involve crimes. This placed the burden of proof on the prosecution to define the specific ways in which the rights of the victims in the cases under trial had been violated, although it was, in general, quite obvious that the manner in which these experimental subjects had been used was only one of a great variety of means by which racial and political prisoners in concentration camps were killed.

In order to define conditions under which medical experimentation on human beings is ethically and legally permissible, I prepared a memorandum entitled “Ethical and Non-Ethical Experimentation on Human Beings,” which was submitted to the United States Chief of Counsel for War Crimes and the Court on April 15, 1947. With additions derived from Dr. Andrew C. Ivy’s testimony of June 12, 13, and 14, 1947, this memorandum became the basis

Annals New York Academy of Sciences


of the so-called Nuremberg Code incorporated in the Judgment.' My original memorandum was subsequently published in its entirety, translated into French, in Commander Bayles' book.3

In searching through the literature for precedent, I had found a most unequivocal statement in the German literature (Ebermayer) .3

The American statutes and court decisions on cases involving the use of new medical or surgical techniques, or new drugs which had not yet attained ortho- dox status, supplied additional leads,"!' as did, of course, the classical guide for medical conduct, the age-old Hippocratic oath.

I proposed six points as essential requirements to render experiments on human beings ethically and legally permissible (see Appendix to this paper).

The judges enlarged these criteria to ten points (see Appendix I at the back of this monograph) by dividing my point No. 4 into three separate points, and by adding two provisions for prompt termination of an experiment at the discretion of the investigator, or at the request of an experimental subject. These were incorporated in their final judgment as the basic principles that must be observed in order to satisfy moral, ethical, and legal concepts with regard to medical experiments. However, they omitted from my original point No. 1 provisions for valid consent in the case of mentally sick subjects to be obtained from the next of kin and from the patient whenever possible, probably because they did not apply in the specific cases under trial. The essentials of these ten conditions required to render experiments in human beings ethical and legal, are the following:

1. Enlightened voluntary consent, in the absence of duress, coercion, or deceit, and with full disclosure of hazards and inconveniences.

2. Humanitarian purpose, unachievable by other means. 3. A sound basis of all other sources of knowledge, including animal

4. Avoidance of all unnecessary physical and mental suffering and/or injury. 5. Absence of n priori reason to believe that death or disabling injury will

occur. 6 . Risk proportionate to humanitarian importance of the problem. 7. Proper preparation for adequate protection against even remote possi-

8. Scientific qualification of the experimenter, who must possess the highest

9. Option for discontinuance always available to subject on request. 10. Discontinuance mandatory whenever experimenter discerns danger of

injury, disability, or death of subject. These ten points constitute what is now known as the Nuremberg Code, a

useful guide setting the limits for experimental research on human beings. It is evident, of course, that the crimes to which this code owes its formulation could not have occurred in any country in which the ordinary laws concerning murder, manslaughter, mayhem, assault, and battery had not been suspended in regard to all or certain groups of its citizens and inhabitants. This code is also unlikely to prevent another dictatorial government from repeating the crimes of the National Socialist Government. Nevertheless, it is a useful measure by which to prevent in less blatant settings the consequences of more subtle degrees of contempt for the rights and dignity of certain classes of human beings such as mental defectives, people presumably dying from incurable illnesses, and people otherwise disenfranchised, including prisoners or other inarticulate public

experi ment at ion.

bilities of injury, disability, or death.

degree of skill, and exercise the highest degree of care.

348 Annals New York Academy of Sciences

charges whose rights might be easily disregarded for the apparently compelling reason of an urgent purpose.

The strict adherence to this code is not only in the interest of the experimental subjects, but also in the interest of the investigator, since he thus gains maximally cooperative and understanding subjects, provided the postulate of enlightened consent is not exaggerated to the point where only research scien- tists would qualify. The need for enlightenment should be interpreted in common-sense terms, indicating what a lay person, in the light of his own knowledge and understanding (which is often considerable, in view of the widespread health information now available through newspapers, periodicals, and lectures to the public), should know in general and practical terms. He should be informed about the benefits the experiment may provide for him or for fellow sufferers from his illness, or for humanity in general, and he should be told about the risks, in terms of potential dangers, discomforts, and time spent or saved, in a completely candid manner. Even placebo controls such as in a double-blind crossover study can be candidly explained without interfering with the unbiased attitude of the patient, as detailed above.

It is always relatively easy to obtain enlightened volunteers suffering from illnesses, the customary treatment of which is either uncomfortable, incon- venient, or considered dangerous, such as the prolonged psychotherapies or the various shock treatments in mental illness, or from illnesses the prognosis of which is considered uniformly disabling or disastrous, such as multiple sclerosis or cancer and the like. The various societies composed of sufferers from this latter group of illnesses have contributed to this cooperative research attitude. This sensible attitude is not only not diminished but is actually enhanced by the investigator’s making the point to correct, in the case of multiple sclerosis patients, for instance, the exaggerated ideas of universal disastrousness of prognosis that these societies tend to impart.

In the past 20 years I have carried out and participated in intensive experimental research on human beings in the fields of major mental illness and multiple sclerosis. Strict adherence to the spirit and letter of the Nuremberg Code at no time hindered or hampered our research; on the contrary, it enabled us to carry on when less good preparation of collaborative enlightened consent could have stopped us at a very crucial point.

The use of prisoners as research subjects poses special problems. While prisoners have collaborated effectively, safely, and without violation of their human rights or dignity in many important research studies such as Goldberger’s famous pellagra experiments and many others, it may nevertheless be doubted whether any consent given by a prisoner, who by definition is restricted in his choices, could ever be regarded as truly voluntary. This question cannot be answered in a generally applicable manner, since the answer depends on the specific conditions, atmosphere, and morale in the particular prison involved. But I am certainly of the opinion that, in the absence of coercion and duress other than the fact of confinement per se and if no promise is made or implied that the condition of his confinement will be changed as a result of his participa- tion in an experiment, a prisoner may make a valid voluntary decision to devote his time and energies to a project for the good of humanity, rather than to relatively meaningless tasks.

The malaria experiments at Stateville Prison in Joliet, Ill., during World War II,lo, l1 demonstrate that ethical experimentation on prisoners can be performed by competent medical investigators in accordance with all the tenets


of the Nuremberg Code. This is, of course, largely and simply due to the fact that they were carried out in a community in which human rights have not been abrogated, even for prisoners.

The greatest advantage of the code is to supply the working scientist with a standard against which he can measure his own proposed experimentation. He can then proceed without the uncertainty that might be present had such a standard not been available to him.

REFERENCES

1. TRIALS OF WAR CRIMINALS BEFORE THE NUREMBERG MILITARY TRIBUNALS UNDER CONTROL COUNCIL LAW No. 10. Superintendent of Documents, U. S. Govt. Printing Office, Washington 25, D.C. Oct. 19'46-April 1949. Vol. I: i-xxviii, 1-1004; Vol. II. i-v, 1-898.

2. ALEXANDER, L. 1947. Ethical and Non-Ethical Experimentation on Human Beings. General Ethical, Medico-Legal and Scientific Considerations. Memo- randum submitted to the War Crimes Court (Military Tribunal No. 1) and the United States Chief of Counsel for War Crimes, April 15. (French Translation in Bayle, Reference 3, 1430-1436).

3. BAYLE, F. 1950. Croix GammCe contre CaducCe: Les Expkriences Humaines en Allemagne pendant la deuxikme guerre mondiale. Berlin and Neustadt (Palatinat) I-XXVII: 1-1521.

4. ALEXANDER, L. 1966. Limitations in experimental research on human beings. Lex et Scientia 3: 8-24. (January-March).

5. EBERMAYER, L. 1930. Der Arzt im Recht, Leipzig (Georg Thieme): 1-287. 6. CASES DETERMINED IN THE FIRST DISTRICT OF THE APPELLATE COURT OF

ILLINOIS DURING THE YEAR 1905. 1906. Gen. no. 11723: E. H. Pratt vs. P. J. Davis. Reports of Cases Determined in the Appellate Court of Illinois

7. E. W. H. 1932. Consent as condition of right to perform surgical operation. 1932. American Law Reports 76: 562-571.

8. SMITH, H. W. 1942. Antecedent Grounds of Liability in the Practice of Surgery. IX: A surgeon may be liable for subjecting a patient to experimental methods of treatment without making a full disclosure of the material facts so that the patient may assume or reject the risk: 263-265. XV: Legal responsibility for human experimentation involving risk to life without compensating social or scientific interest: 286-293. Rocky Mountain Law Review 14: 233-293.

9. ARTHUR, W. R. 1945. Some liabilities of the physician in the use of drugs. Rocky Mountain Law Review 17: 131-162; 144-146.

10. HARRISON, R. W. 1946. Dean of the University of Chicago School of Medicine and Biological Sciences: Report on the Malaria Experiments at Stateville Penitentiary in Joliet, Ill.

11. LADIMER, I., & R. NEWMAN. 1963. Clinical Investigation in Medicine. Legal, Ethical and Moral Aspects. Boston University Law-Medicine Research Institute. I-XXIII: 1-5 17.

118: 161-184.

APPENDIX

Conditions rendering Experimentation on Human Beings Ethical and Permissible

1. Legally valid voluntary consent of the experimental subject is essential. This requires specifically:

a. The absence of duress;

350

b. Sufficient disclosure on the part of the experimenter and sufficient understanding on the part of the experimental subject of the exact nature and consequences of the experiment for which he volunteers to permit an enlightened consent.

In the case of mentally ill patients, for the purpose of experiments concerning the nature and treatment of nervous and mental illness or related subjects, such consent of the next of kin or legal guardian is required; whenever the mental state of the patient permits (that is, in those mentally ill patients who are not delirious or confused), his own consent should be obtained in addition.

2. The nature and purpose of the experiment must be humanitarian, with the ultimate aim to cure, treat, or prevent illness, and not concerned with methods of killing or sterilization (ktenology). The motive and purpose of the experiment shall also not be personal or otherwise ulterior.

3. No experiment is permissible if the foregone conclusion exists, or the probability or the a priori reason to believe that death or disabling injury of the experimental subject will occur.

4. Adequate preparations must be made and proper facilities be provided to aid the experimental subject against any remote chance of injury, disability, or death. This provision specifically requires that the degree of skill of all those who are taking an active part as experimenters, and the degree of care which they exercise during the experiment, must be significantly higher than the skill that is considered qualifying and the care that is considered adequate for the performance of standardized medical or surgical procedures, and for the administration of well-established drugs. American courts are very stringent in requiring for the permissible use of any new or unusual technique or drug, irrespective of whether this use is experimental or purely therapeutic, a degree of skill and care on the part of the responsible physician, which is higher than that required for the purpose of routine medical or surgical procedures.

5 . The degree of risk taken should never exceed that determined by the humanitarian importance of the problem to be solved by the experiment. I t is ethically permissible for an experimenter to perform experiments involving significant risks only if the solution, after thorough exploration along all other lines of scientific investigation, is not accessible by any other means, and if he considers the solution of the problem important enough to risk his own life along with the lives of his nonscientific colleagues, such as was done in the case of Walter Reed’s yellow fever experiments.

6. The experiment to be performed must be so designed and based upon the results of thorough thinking-through, investigation of simple physicochemical systems, and of animal experimentation, that the anticipated results will justify the pefformance of an experiment. That is, the experiment must be such as to yield decisive results for the good of society and should not be random and unnecessary in nature.

Annals New York Academy of Sciences