Human experimentation in historical and ethical perspectives

SW.

SC;. Mrd

Prmted

Vol.

m Great

16. pp.

1429 10 1448.

0277-9536’82.

1982

HUMAN

15132Y-?0$03.00.0 Pergamon

Entam

EXPERIMENTATION IN HISTORICAL ETHICAL PERSPECTIVES*

Press

Ltd

AND

NORMAN HOWARD-JONES 38. Chemin

Colladon.

1209 Geneva.

Switzerland

Abstract-During the past two decades there has been an extraordinary proliferation of articles. books, reports. laws and regulations. and meetings concerned with the ethical aspects of biomedical interventions on human beings and. in particular. those of an experimental nature. Such writings, the vast majority of which are in the English language, necessarily tend to become more and more repetitive in the sense that they bring to light no new facts. or even ideas. but are variations on the same themes according to the cultural and professional backgrounds of their authors. For some of the older problems of medical ethics. particularly those linked to human reproductionsuch as induced abortion. sterilization. and artificial insemination from a donor-it would be illusory to expect a consensus at the national. far less the international. level, for religious dogma or deeply-felt personal convictions may render discussion sterile. However. for the increasingly important subject of biomedical experimentation involving human subjects it might well be possible to arrive at an international consensus. as has already been done in very broad terms by the World Medical Association in the Tokyo (1975) revision of the Declaration of Helsinki (Helsinki II). It was m the belief that a more comprehensive consensus might be achieved that the World Health Organization and the Council for International Organizations of Medical Sciences embarked in 1978 on a joint prqiect for the establishment of international etlucal guidelines for biomedical research involving human subjects. taking Helsinki II as their starting point. In some of the existing literature on bioethics there are brief historical introductions. but these are usualI> very rudimental). It was therefore decided that. in order to place the WHOKIOMS project in the broadest possible perspective. an attempt should be made to provide a further. if not comprehensive, account of the historical aspects of human experimentation and of the circumstances in which formal ethical codes evolved.

HUMAN

EXPERIMENTATION

IN

GENERAL

Up to the end of the eighteenth century. the practice of medicine had hardly advanced since the time of the Hippocratic writings-a period spanning almost two and a half millenia. There were some outstanding-if isolated-advances in anatomy, physiology and pathology. but these had no useful applications in medical practice. Most surgical procedures-such as the setting of fractures. the extraction of bullets or other foreign bodies. the dresses of wounds and lithotomyhad been known since time imtnemorial and before either anaesthesia or antisepsis had been introduced surgery could do little to enlarge its dimensions. Although a fe\s active medicinal preparations were known. such as opium. cinchona bark and digitalis. ignorance of their modes of action prevented their rational use. The /~~reria r~~edica consisted of almost every known animal. vegetable. or mineral substance and. to these. therapeutic properties were arbitrarily assigned. In fact. all pharmacotherapy was experimental in the sense that it was empirical rather than based on scientific reasoning. In his “Introduction to the Study of Experimental Medicine”. first published in 1865. Claude Bernard referred to several examples of medical experiments *Amplified version

of a stud! prepared as a background document for a Joint project by the World Health Organization and the Council for International Organizattons of Medical Sciences for the establishment of international ethical guidelines for biomedical research mvolvlng human sublects.

made on condemned criminals in ancient times [I], but these are probably as much in the realm of legend as of history. A particularly persistent legend is that human vivisection was practised in Ptolemaic Alexandria and both ancient and modern historians have expressed contrary opinions as to whether this legend had a factual basis. However, as stated in a recent and erudite study of this question, such opinions are. at their best, ‘learned guesses’ [Z]. A much more recent and unquestionably authentic exampie of a medical experiment on human beings took place in 1721. The wife of the British Ambassador to Turkey, Lady Mary Wortley Montagu. had returned to England in 1718, where she successfully promoted the use of variolation. Through her friend. the Princess Caroline, she persuaded George I to offer a free pardon to any condemned inmate of Newgate Prison who would agree to be variolated [3]. Six of them-3 men and 3 women. ages ranging from 19 to 36 years-volunteered. and on 9 August 1721 they were all inoculated with supposedly infectious smallpox matter by Lady Mary’s surgeon-Charles Maitland. However. the expected reactions did not appear and 3 days later they were successfully inoculated with fresh material. They developed conspicuous local lesions. but had little systemic disturbance. On 6 September they were all. in Maitland’s words. “dismiss’d to their several Counties and Habitations’* [4]. This experiment may be viewed from two different aspects. On the one hand. it was unique in the sense that never could consent to participate in an experiment have been more genuinely informed. Smallpox I429

1430

NORMAN

HOWARD-JONES

was then a virtually universal disease in Europe and. except for rural areas of low population density. few escaped an attack at some time in their lives. For all the Newgate volunteers. the effects of smallpox were therefore a matter of everyday experience. On the other hand. the view is widely. although not universally. held that prisoners and some other special groups cannot give valid consent because they are influenced by pecuniary rewards or other incentives. or by moral suasion. In the case of Maitland’s experiment. the options for the prisoners were: volunteer or be hanged! A stronger incentive to Volunteer could hardly be imagined-but. does this make the experiment unethical‘? If not. the question arises: where is the line to be drawn between volunteering to avoid the hangman’s noose and volunteering in the hope. whether justified or not. of improved prison conditions and earlier parole’? If this early example of human experimentation is vulnerable to criticism. it is surely not on ethical grounds but because the results were inconclusive. In fact, available evidence points to variolation as the cause of many deaths and the source of many local outbreaks of smallpox. The survival of all of these six volunteers offered no evidence that others would be equally fortunate. On 14 May 1796 Edward Jenner performed his crucial experiment of inoculating a boy of about 8 years’ old with cowpox matter, 7 weeks later inoculating the boy with pus taken directly from a smallpox lesion without producing an attack of th5 disease [S]. By today’s standards. Jenner’s use of a child for his experiment would be considered ethically reprehensiblc:. especially as a consenting adult would have served his purpose equally well. However, in his favour it should be pointed out that he was thoroughly familiar with the manifestations of accidental cowpox infections in man and knew them to be normally mild and to afford apparent protection against smallpox. The two immunological experiments described above were isolated examples of experimentation on human subjects. But. during the nineteenth century almost all medical practice. with the exception of some surgical procedures, continued to be experimental in the sense that it was empirical. The most heroic, dangerous-and even lethal--procedures were undertaken, not as experiments, but because it was believed that they would save lives. Thus, when an entirely new disease-asiatic cholera-invaded Europe at the beginning of the 1830s. the most generally accepted treatment for it for almost a decade was blood-letting. Desperate attempts were made to obtain blood. even though it was of the constency and colour of tar, from patients dying of dehydration. in the mistaken belief that the heart was embarassed by a centripetal flow of the blood. In 1831 a German surgeon. J. F. Dieffenbach, on the insistence of the doctors of the Berlin Cholera Hospital that it was therapeutically indispensable to obtain blood. introduced a catheter into the left ventricle of the heart of a moribund cholera patient [6]. Nevertheless. his publication won him the Prix Montyon of the Paris Academy of Sciences. Other curative measures solemnly applied and advocated-not as experiments but as preferred treat-

ments for cholera-were pouring boiline water on the umbilical region. electric shocks. plugiing the ‘tnus with a cork retained by a T-bandage and the administration of castor oil to patients dymg of the effects of diarrhoea. The most fervent protagonist of the last of these treatments was Sir George Johnson. who became Physician Extraordinary to Queen Victoria. and who for 42 years never ceased to maintain that castor oil was the sovereign remedy for cholera. As he unblushingly admitted. he was widely known as “Castor Oil Johnson” [‘il. Throughout the nineteenth century. most medical experiments on human subjects-as opposed to empirical but supposedly beneficial interventionsappear to have been made by physicians upon themselves. These are discussed later under the heading ‘Auto-experiments’. English law on therapeutic experiments by doctors on their patients in Dieffenbach’s time was summarized by Willcock [S] in the following terms: “By experiments we are not to be understood as speakmg of the wild and dangerous practices of rash and Ignorant practitioners, which fall properly under the head of want of skill and knowledge; but of dellberate acts of men of considerable knowledge and undoubted talent, differing from those prescribed by the ordinary rules of practice. but which they have good reason. from a knowledge of the human system and of the particular disease.-and if the experiment be medical as opposed to surgical. of the nature of the medicine.-to believe will be attended with benelit to the patient. although the novelty of the undertakmg does not leave the result altogether free of doubt”

Willcock continued

(italics not m original):

“When an experiment (‘OIlSPHI Of‘ tl1r purry

of this kind is performed ~cirh tier s1rhjrctrd 10 II. tr/w,’ iw hs htwl iujiirmed rhut ir is au euprritnrnt. the practitioner is answerable neither in damages to the individual. nor on a criminal proceeding;. But if the practitioner performs his experiment wthmrr yivrmq .swh infiwtnntim to. rd ohtaiuit~y the consenr of; his purirnt. he is liable to compensate m damages any injury which may arise from his adopting a new method of treatment”.

Thus was enunciated. exactly 150 y,ears ago. a formulation of the essentials of the principle of informed consent although. as is the case today. the onus of making a decision was borne by the least qualified to decide: the subject. Herein lies a major weakness of the concept of informed consent. for the subject is ‘informed’ in the sense that he has been fold. but not in the sense that he understands well enough to make a reasoned decision in full cognizance of all implications of the experiment. Thomas Percival’s well-known Medical Ethics (1803) is sometimes cited as if it were a sort of bioethical milestone, but it contains nothing of relevance to human experimentation and is concerned largely with such matters as the deportment of physicians L?S-his their patients and each other, ‘pecuniary acknowledgement’ by patients. and the duty of physicians to attend divine service on the Sabbath. One would hardly suppose from this pious exegesis that the medical (as opposed to surgical) practice of Percival’s time was. had always been and was to remain for some years, an elaborate hoax. fully justifying the dictum of George Bernard Shaw that “all professions are conspiracies against the laity”.

Human experimentation

in historical and ethical perspectives

Claude Bernard would have endorsed Shaw’s dictum, for he had no illusions about the scientific standing of the medical practice of his own time. In 1867 he recalled that in his inaugural lecture as professor of medicine at the College de France in 1847 he had opened

with

the following

statement

[9]:

“Scientific medicine. which I have the responsibility to teach. does not exist. We can only prepare the materials for future generations by founding and developing experimental physiology, which must serve as the basis for experimental medicine”.

For most practising doctors, Bernard said, medicine was nothing but an industry, the object of which was to have as many patients as possible and draw the maximum possible profit from them. He cites “a renowned physician of the [Paris] School of Medicine” as insisting that patients asked for nothing better than to be deceived and that if they were not treated they would go elsewhere. Bernard adds the devastating comment that “Molitre’s consultations are as valid today as in his time, except that the words have changed” [lo]. Referring to experimental medicine. Bernard postulated in 1865 that the “principle of medical and surgical morality” was never to perform on a human subject an experiment whose outcome could only be harmful in some degree. even though the result could be very interesting scientifically, and therefore of interest for the health of others [I I]. He approved investigations made on the bodies of criminals immediately after decapitation. and also the administration by the helminthologist. C. J. Davaine, of larvae of intestinal worms to a condemned woman with a view to postmortem examination of her entrails. He also endorsed experiments on dying patients if they caused no suffering. Bernard concluded with the formulation: “experiments that can only have harmful results are forbidden: those that are harmless are permitted: those that can do good are obligatory”. The last clause of this formulation would seem to beg the question. because if it is known in advance that a treatment can do good it is not an experiment. Moreover. medical histor) is strewn with examples of interventions thought to be beneficial that were in fact noxious or even lethal, of which blood-letting as a treatment for cholera is an outstanding example. Bernard himself, deploring the posr hoc ergo propter hoc fallacy. further on pointed out that it was only by the use of untreated controls that the treatment of pneumonia by venesection. previously believed to be “very efficacious”, had been shown to be a “therapeutic illusion” [ 121. ALTO-EXPERIMENTS The ‘nineteenth centur) was characterized by numerous experiments by physicians upon themselves. A pioneer of auto-experimentation was Johann Christian Gottfried Jiirg (1779-18561, who in 1825 published a volume of over 500 pages on “Contributions to a Future Pharmacology by Trials of Drugs on Health! Humans” [ 131. J&-g pointed out that many had published accounts of the effects of drugs upon themselves. but that such observations \vere inconclusive in that they bore the

1431

stamp of the individual experimenter, and he deplored the fallacy of conclusions drawn from the pooling of individual reports based upon differing conditions of dosage and subject. He was convinced that the way to rational therapeutics was by detailed collective observations, under uniform conditions, on the effects of drugs on healthy subjects. For this purpose. he constituted an ‘experimenting society’ (experimmtirende Gesellschajl) of 27 volunteers, including himself, who ingested progressively increasing doses of 17 different drugs, noting in detail date, time, dose, and the sensations experienced. This laudable initiative produced no lasting results except, perhaps, the conclusion that valerian, then and for many years afterwards reputedly useful for the treatment of hysteria, was judged by Jiirg to be “an unreliable drug.. without any discernible action”. Early in the nineteenth century, self-infections. sometimes fatal, by anticontagioist physicians were not uncommon. and the voluminous early literature on cholera in the 1830s contains numerous descriptions of auto-experiments, including the tasting of vomitus and dejections of cholera patients. deliberately remaining in close contact with them, or sleeping in the beds of those who had succumbed to the disease. After Robert Koch’s definitive incrimination in 1884 of the vibrio as the cholera pathogen, many attempts to refute his findings were made by physicians who had swallowed living cultures of the organism. By 1894. Drasche was able to cite 27 examples of such experiments [ 143. The best-known example is that of the then Professor Hygiene at Munich, Max von Pettenkofer, who on 7 October 1982 first neutralized his gastric juices with sodium bicarbonate and then swallowed a broth culture containing an estimated thousand million vibrios [15], suffering as a consequence only a mild diarrhoea. Numerous examples could be cited of auto-experiments with other living pathogens [ 161. Many auto-experiments were made in fields other than that of communicable diseases. For example, on the evening of 4 November 1847, the obstetrician, James Young Simpson and two of his medical colleagues decided, in an attempt to find an anaesthetic agent superior to ether, to inhale chloroform simultaneously. regaining consciousness only to find themselves prostrate on the floor but unharmed [ 173. Very soon after this experiment Simpson introduced the use of chloroform in his obstetric practice. It was to be many years before the dangers of this anaesthetic agent were recognized. An auto-experiment of another kind and of relatively recent times is that of the neurologist. Henry Head. who in 1903 severed some of his own sensory nerves in order to study the distribution of loss of sensation and the stages of its restoration [IS]. One of the most remarkable auto-experiments of all times was that of Werner Forssmann. who in 1929. after having perfected his technique on cadavers. used himself as the first living human subJect for the passage of a catheter into the right ventricle of the heart. verifying its position by radiography. for which purpose he walked some distance and up some stairs. with the catheter in siru. thus demonstrating that this procedure could be performed safely and without discomfort [19]. For this courageous experiment and

NORMAN

1432

subsequent observations. Forssman of the Nobel Prize in 1956. CLINICAL

TRIALS

OF

HOWARD-JONES

was a co-recipient

DRL;GS

By far the most common form of experimentation on human subjects is the administration of new drugs. or of older drugs for new indications. The overwhelming impression left by a consideration of the experience of the past is that during the milleniai history of medical practice it is only almost within living memory that drug therapy-which was virtually synonymous with internal medicine-became, with very few exceptions. anything more than a refined form of charlatanism. It is therefore relevant to reflect on the conditions under which the pharmacotherapeutic revolution took place. and to note that as pharmacotherapy and immunotherapy have become ever more effective in the treatment and prevention of disease. new risks and the increasing intervention of official regulatory agencies have entailed a parallel increase of progressively burdensome regulatory measures. Although the nineteenth century witnessed remarkable medical advances in many directions, few of these had a direct bearing on the rational treatment of disease. Many active principles of crude drugs were isolated in pure form, such as atropine, cocaine, morphine and strychnine and new organic compounds such as chloroform and phenol were synthesized, all of which compounds but strychnine were to have valuable applications in surgery for some time after their discovery. An accidental finding that was to have epoch-making consequences for medicine was that of the aniline dye mauve. which William Henry Perkin produced in an attempt to synthesize quinine. This compound “opened the way for the dyestuff industry” [ZO], which was subsequently to diversify into pharmaceutical preparations. thus laying the technological infrastructure for modern pharmacotherapeutics. The nineteenth century also saw the development of diagnostic instruments such as auriscopes, clinical thermometers, haemoglobinometers, ophthalmoscopes. sphygmomanometers, stethoscopes and X-rays, as also biochemical tests of bodily function. The birth of medical bacteriology, while throwing a flood of light on the hitherto obscure nature of communicable diseases. was as yet of little help except in their diagnosis. While anaesthesia and antisepsis made possible considerable progress in surgical treatments. and there was great progress in some of the more theoretical aspects of internal medicine, pharmacotherapy stagnated for most of the century. A few enlightened physicians were fully aware of the inadequacies of their therapeutic armamentarium. Outstanding among these was Oliver Wendell Homes of Boston. who had no illusions about the futility of the drug therapy of his own time. On 6 November 1861 he told the medical class of Harvard University [21]:

“the disgrace of medicine has been that colossal system of self-dcccption. m obtdience to which mines have been emptied of their cankering mrnerals, the vegetable kingdom robbed of all its noxious growths, the entrails of animals taxed for their Impurities. the poison-bags of reptiles

drained of their venom. and all the inconceivable tions thus obtalned thrust down the throats beings .“.

abomlnnof human

In the previous year. Holmes had declared to the Massachusetts Medical Society that much of the blame for this indiscriminate medication rested “with the public itself. which insists on being poisoned” [22]. and further on he stated that: “if the whole materla medica. OS UON IIS&. could be to the bottom of the sea. it would be all the better for mankmd

and all the worse for the fishes” [Z3]. In 1913, slightly more than half a century later. the venerable Sir Thomas Clifford Allbutt expressed himself in similarly sceptical terms about his own experiences of medical practice in the Victorian era [24]: “When I began practice

it was customark at every consultation to prepare a writmg-table. pens and ink for ‘the prescription’. This script. wasof formidable compositlon. a drug for every symptom. ,md a few more for the pool: it was solemnly set forth dnd slgncd b) two or more phys-

icians: by the patlent’s friends this was understood. long enough. to be the organ of his restoration”.

If but

A notable departure from prescribing of Irrational medicinal mixtures was the introduction in 1867 by Sir Lauder Brunton of amyl nitrite by inhalation for the relief of angina pectoris. Forty years later Brunton claimed that this discovery could “fairly be regarded as the Hurst complete example therapeutics based on experlmentai pathological macological data” [Xl.

of ratIonal and phar-

An advance of tremendous significance for pharmacotherapy was inaugurated during the last two decades of the nineteenth century, when the chemical industry started to produce synthetic pharmaceuticals, such as Antipyrine (phenazone), aspirin. phenacetin, Pyramidon (amidopyrine) and sulphonal. The twentieth century was to see the gradual replacement of the traditional individual. polypharmaceutical, prescriptions by single substances with specific effect. manufactured and mostly originated by the pharmaceutical industry. Almost 40 years ago, Sir Henry Dale wrote: “the old-fashioned dispensing of mixtures to Individual prescriptions will doubtless linger for a time: but I think that we must regard it as inevitable that the preparation of the remedies required by a progressive therapeutics will eventually fall entirely into the hands of a sclentlfic. largescale manufacture. and that the role of the individual pharmacist in relation to them will become little more than an intelligent retail distribution of ready-made and centrally standardized products ” [26].

Dale’s prediction has been amply fultilled. although certain sectors of the pharmaceutical industry still advertize and purvey through the medical profession irrational mixture under proprietary names. For these. as for more valid products of the Industry. the pharmacy has become the retailer for industry, while the individual physician is the vicarious customer. Supposed remedies advertized to and bought directly by the general public are legion and fraudulent claims for them are made with impunity in the mass media. The public still-in the words of Oliver Wendell Holmes-“insists on being poisoned”.

Human

experimentation

While the older. naturally-occurring, medicaments of plant or mineral origin were mostly without demonstrable therapeutic action, long experience had shown that. in the doses used. they caused no serious harm. The multiplication of new wholly or partially synthetic therapeutic substances that had never existed before. and that sometimes had undesirable or even disastrous secondary effects, gave rise to a growing recognition of the need for extreme caution in recognizing them as suitable for routine use in clinical practice. THE

PHARMACOTHERAPEUTIC

REVOLUTION

Very early in the twentieth century the foundations for the chemotherapy of microbial infections were laid. and by 1906 the trypanostatic effects of atoxyl and trypan-blue had been demonstrated. In the following year Paul Ehrlich showed a similar effect for trypan-red. He not only coined the term ‘chemotherapy’ but also formulated its theoretical basis. In a lecture to the Berlin Medical Society on 13 February 1907. Ehrlich maintained [27]: “What we want is a chrn~orhc,rupitr sprcijiccr. that is, we are looking for chemical agents which. on the one hand, are taken up by certain parasites and are able to kill them and, on the other hand. in the quantities necessary for this lethal action. are tolerated by the organism without too great damage”. He expanded this concept by referring to the properties of chemotherapeutic drugs as etiotropic and organotropic. Only if etiotropy were stronger than organotropy could such a drug be regarded as specific for the infection. Ehrlich clearly accepted that drugs might have harmful effects, provided that these were outweighed by their beneficial action. In 1908, he shared the Nobel Prize for Medicine with Elie Metchnikoff. and in 1910 he introduced, with Sahachiro Hata. Salvarsan (arsphenamine) in a monograph on the chemotherapy of the ‘spirilloses’ in which he included syphilis, relapsing fever. avian spirochaetosis and yaws. During the first 20 years after the death of Ehrlich in 191.5 there was. after a pause during the First World War. an extraordinary proliferation of synthetic or semi-synthetic pharmaceutical products, which progressively displaced the traditional medicines. Very fe& of these were aimed at the causes of disease-or. in Ehrlich’s terminology. were etiotropit-with the notable exceptions of synthetic antimalarials and trypanostatic drugs. New synthetic drugs included general and local anaesthetics. analeptics, analgesics. antispasmodics. amphetamines. mercurial diuretics. hypnotics and sedatives. Insulin saved the lives of diabetics. commercially available vitamins multiplied. and partially synthesized sex steroids were soon to become available. Immunoprophylaxis made relativel) little progress. During this period, the pharmaceutical industr) became the indispensable ally of the medical profession.

A NE\\

ERA IN CHEMOTHERAPI

With the publication on 15 February 1935 of Gerhard Domagk’s classical paper on the treatment of

in historical

and ethical

perspectives

I423

experimental infections in mice by Prontosit. for which he was awarded the Nobel Prize in 1939, chemotherapy entered a new era [28]. Domagk opened his paper by pointing out that hitherto only parasitic and spirochaetal infections had proved to be susceptible to chemotherapy, enumerating Atebrin (mepacrine) and Plasmochin (pamaquine) for malaria. Germanin (suramine) for trypanosomiasis, Neostibosan (stibosamine) for leishmaniasis and Salvarsan (arsphenamine) for syphilis. It is noteworthy that Prontosil and the five predecessors mentioned by Domagk had all been originated in the same country, in which there was a fruitful collaboration between medical scientists and the thriving chemical industry. In his paper Domagk presented a table showing that of a total of 26 mice infected with haemolytic streptococci from a human patient. 14 that were untreated all died. The remaining 12. treated with Prontosil orally, all survived, although for how long was not stated. In spite of the small number of mice involved, this experiment was conclusive. Other workers were to obtain similar, if not so conclusive, results. Curiously enough Domagk’s table shows that the experiment covered the period 2&28 December 1933-almost 14 months before its publication. In the meantime, samples of Prontosil had been made available for limited clinical trials to individual German physicians. originally under the name Streptozon, and reports of two of these describing results in a variety of streptococcal infections followed Domagk’s paper. This was the first time that a drug had been shown to be effective against any bacterial infection, and especially against one of the most common and most dangerous of them. but the significance of these findings seems not to have been as widely appreciated as might have been expected. In 1936 were published, under the auspices of the Therapeutic Trials’ Committee of the Medical Research Council of the United Kingdom, the results of a clinical trial of Prontosil in the treatment of human puerperal infections with haemolytic streptococci [29]. As a preliminary to this trial, Domagk’s results with experimental streptococcal infections of mice were largely confirmed. To have withheld, while awaiting the outcome of the deliberations of an ethical review committee, a drug of proven efficacity against the haemolytic streptococcus in riroalthough not in uirro-from women whose lives were in peril, and for whom there was no other effective treatment, would have been more ‘ethical’ than reasonable. The favourable results of this trial, conducted under the auspices of a body of incontestable scientific authority, aroused wide interest. AS P.H. Long and E. A. Bliss wrote a few years later [30] “One can say without fear of contradiction or thought of disparagement of the observations made by the continental investigators that these original reports.. aroused the interest of the English-speaking world upon the subject of bacterial chemotherapy”. The lack of in vitro activity of Prontosil against streptococci at first confused understanding of the mo,de of its action. It was supposed that it might not

NORMAN HOWARD-JONES

I-l?4

act directly upon the pathogens but in some way stimulate immunological mechanisms of the host. Domagk himself. in his original report. stated that Prontosil “as a true chemotherapeutical acts only in the living organism” [31]. The answer to this riddle was provided by workers at the Institut Pasteur. Paris [32]. Prontosil was a red dye. part of its chemical structure consisting of sulfanilamide. The Paris workers found that this chemically simpler component was not only as strongly bacteriostatic as Prontosil, but also less toxic. It was later shown that sulfanilamide was also active in vitro. However. this compound had been synthesized as long ago as 1908 by P. Gelmo, and in 1919 M. Heidelberger and W. A. Jacobs had found that a compound of it was “highly bactericidal in citro” [33]. Sulfanilamide was therefore not patentable, and this implies a possible ethical question: Did the manufacturers deliberately place on the market an unnecessarily complex compound in order to have a patentable product’?

THE

ELIXIR

DISASTER

Sulfanilamide rapidly became established as an invaluable antistreptococcal drug, without apparent danger, and was followed by 1938 by sulfapyridine (active against pneumococci) and in 1940 by sulfathiarole (staphylococci) and sulfaguanidine (intestinal infections). In the latter year. H. W. Florey’s group published the first observations of the treatment with penicillin of experimental infections in mice [34] and in 1941 they reported the first results of clinical trials in IO cases [35]. In the short space of 7 years, the treatment of bacterial infections had been revolutionized. In 1937 an obscure pharmaceutical firm in the U.S.A. placed on the market a solution, which it had tested only for ‘flavor. appearance, and fragrance’, described as ‘Elixir of Sulfanilamide’. There were soon alarming reports of toxic reactions, and before the preparation was withdrawn it had been directly responsible for over 100 deaths [36]. Investigations initiated by the American Medical Association and others showed that the solventdiethylene glycol-was the sole cause of the deaths and of non-fatal poisoning [37]. This disaster aroused nation-wide concern and in the period October 1937-May 1938 the Journal of the American Medicul Associution published six editorials dealing with it, that for 20 November stating [38]: “Seldom has any catastrophe stirred the United States to the extent to which press and public have been aroused by the needless deaths resulting from the Elixir of Sulfanil-

amide-Massenger”. Responding to resolutions adopted by the Senate and the House of Representatives, the Secretary of Agriculture submitted a report to Congress on 26 November 1937 [39]. In this report he proposed that: new drugs should be generally available only after licensing on the basis of “experimental and clinical tests”; drugs dangerous to health should be prohibited; drug labels should contain directions for use and a full declaration of their composition.

New drugs were defined as drugs that had not been used sufficiently to become generally recognized us safe: combinations. not general recognized as safe. of well-known drugs: drugs or drug combinations labelled for use in higher dosage than usual. In 1938 and as a direct consequence of the ‘elisir’ mishap. the provisions proposed by the U.S. Department of Agriculture were incorporated. in the face of strong opposition by business interests and in rather attenuated form. in the Federal Food. Drug and Cosmetic Act. Previously. the only restrictions on the marketing of drugs were that they should not be adulterated or misbranded or the subject of false therapeutic claims, The new Act brought under the control of the Food and Drug Administration any new drug or device for the diagnosis. treatment. or prevention of disease in man or animals. and prohibited tratfic in them unless they had. by the criteria of the time. been shown to be safe. There were additional provisions as to labelling. and an exemption was made for‘drugs intended solely for investigation by qualified scientists. These federal regulatory requirements had as their object the protection of the public from irresponsible or unscrupulous manufacturers of alleged remedies. rather than from experiments by medical investigators. Some three decades later much more stringent regulations were enacted. These. far from proscribing human experimentation. made it obligatory. It is to be noted that these regulations apply only to research projects supported by federal funds. Nevertheless. they have been widely accepted in the United States as being of general applicability.

THE THALIDOMIDE

TRAGEDY

Thalidomide, first synthesized in 1956. was bY 1958 widely used in medical practice as a sedative. It had the unusual property of being well tolerated in doses many times higher than those required for sedation. Several patients who had ingested large overdoses suffered brief periods of unconsciousness followed by complete recovery and dogs given 1000 times the sedative dose suffered no ill effects [40]. With this exceptionally wide safety margin. thalidomide appeared to be an ideal sedative for pregnant women and it was not until late in 1961 that the first report linking it with congenital defects of the newborn was published. This was a complication without precedent, and there was no means by which it could have been foreseen, except by tests of teratogenicity in animals, which were not at the time considered to be necessary, and for which, indeed. there seemed to be no obvious reason. It is safe to say that. at the time that thalidomide was introduced into medical practice. any regulatory body would have passed it as an exceptionally safe sedative. In spite of the unprecedented and unpredictable. nature of this disaster, it provided a powerful incentive to a tightening of regulatory requirements. as had the elixir of sulfanilamide a quarter of a century before, and among the investigations of new drugs required by U.S. Federal Regulations are tests for teratogenicity in several animal species [4l].

Human BEFORE AND AFTER THE

experimentation

in historical

NUREMBERC

TRIALS

For much of the nineteenth and early twentieth centuries German medical science was second to none. Germany was also the cradle of the modern pharmaceutical industry. of which it had a virtual monopoly for many years. It was also in the same country that. as early as 1900. a ministerial directive was issued laying down basic ethical criteria for medical interventions other than for diagnostic. therapeutic, or prophylactic purposes (presumably those already in current use) [42]. This directive, addressed to heads of clinics, polyclinics, and other institutions for the sick. appeared over the signature of the Prussian Minister for Spiritual etc. (sic] Affairs. It provided that such interventions should in no circumstances be undertaken on minors or those incompetent for other reasons. or on other persons unless the) had given their unequivocal consent in the knowledge of any possible detrimental consequences. Further. such interventions might be undertaken only by heads of clinics or other institutions. and must be recorded on the patients’ case-sheets. Ten years later. Paul Ehrlich discussed in some detail the ethical implications of the use of new drugs on human subjects. and his conclusions may be summarized as follows [43]: often mamtained that for the treatment of human beings. and especially in the case of syphilis. only absolutely harmless substances should be used. If this restriction is accepted. any progress in chemotherapy becomes im-

“It is

possible”, Ehrlich then poses the question: What would be the position of the surgeon who, faced with a condition that did not threaten life. could only perform an operation that was absolutely devoid of danger‘? Clearly. he says. there can be no difference between therapeutic “ethics” (Morul) of internists and of surgeons. He concludes that, just as the surgeon carefully weighs the threat of the pathological condition against the statistical risks of the operation. and gicr rhe putirr~ ,firl/ inforrlluriorl on the risk-benefit ratio (Gqfiil2~d~trlClscluorirrlrierl). so must the chemotherapeutist be able to tell his parienr whar possibilities of risk exisr and how frequent rhq, are. Ehrlich adds that it is not bossible to give such information accurately until a sufficiently large casematerial from a sufficient number of centres is available. He thus implicitI, concedes that there can be no guarantee that first trials on human subjects will not be attended by unexpected risks that the most exhaustive animal trials have failed to reveal. He suggests no solution to this dilemma. for which there is not-and by its very nature never can be-a solution. After 70 years. there is not really much to add to Ehrlich’s formulation. PL’BLIC

CONCERN

ABOUT

HUMAN

EXPERlMENTATlO?i

In the 1920s in Germany. criticism in the daily press. as also in the Reichstag. of alleged unethical conduct on the part of the medical profession reached a crescendo unparalled in any other country. Early in 1930 a spokesman of the National Health Administration (Reichsgescrildlteirsar,lr) listed some of the

and ethical

accusations

perspectives

1435

made as follows [44]:

“Naked cynicism: placing the lives of small children on the same level as those of experimental animals (rats); dubious experiments having no therapeutic purpose: science sailing under false colours: crimes against the health of defenceless children; lack of sensibility; mental and physical torture; martyrization of children in hospitals; the worst forms of charlatanism: disgustingly shameful abominations in the name of science run mad; horrors of the darkest middle

ages, outstripping

the infamous

and the hangman;

social

injustice:

deeds of the Inquisition discrimination between

the rich and the poor”.

Even such extreme language was to be fully vindicated little more than a decade later in respect of the acts of certain German physicians, but it seems hardly probable that it was justified at the time. Miiller (see below) referred in 1930 to the immense number of new medicaments “thrown onto the market and advertised”. and to accusations that hospitals were working for the chemical industry, and hence for big business. Here. in the fertility of the German pharmaceutical industry, may be sought a partial explanation for the avalanche of public criticisms of the medical profession, which was perhaps not sufficiently critical in exploiting the multiplicity of new remedies placed at its disposal. Moreover, the strength of scientific medicine and of the pharmaceutical industry in Germany had a parallel in the strength of cultist parallel systems, which constituted a natural antimedical lobby. Be that as it may, the Reichsgesundheirsumr took these attacks seriously, and on 14 March 1930 the National Health Council (Reichsgesundheir.srur) held a session to discuss the “permissibility of medical experiments on healthy and sick human subjects”, the two speakers being Friedrich Miiller of Munich and Alfons Stauder of Nuremberg and their contributions totalling approx. 10.000 words. Miiller expounded familiar arguments on the necessity of human experimentation, citing examples of its benefits. and dismissing the popular idea that it was carried out only on patients in wards of the 3rd class [45]. In fact, he said. his own experiments on the insulin treatment of diabetes and the liver treatment of pernicious anaemia had been performed on patients in private rooms because of their “greater intelligence”. Miiller then postulated some principles that should govern human experimentation: Experiments of unknown outcome (mir unbekannren Ausgang) should be undertaken only with the agreement of the patient, except in the case of confused or unconscious patients in emergencies. This condition, he said, was SO selfevident that no further explanation was necessary, but it is in fact a meaningless statement, for any experiment has, by definition. an unknown outcome. Other conditions were that such experiments should be carefully planned, their possible consequences carefully weighed, and they should be prudently carried out. They should not be entrusted to inexperienced staff. and the chief physician should assume entire responsibility for them. Miiiler concluded a rather rambling discourse by unconditionally repudiating a- suggestion. made during discussions of the Berlin Medical Board (Berliner Aerztekammer) and taken up by the general press, that

1436

NORMAX HOWARD-JWES

there should be an official regulatory body to which proposals for experiments on man should be submitted. Such a measure. he said, would result in the most serious conflicts. would paralyse the creativity of hospital physicians. and would place the scientific activities of German hospitals at a disadvantage by comparison with hospitals of other countries. Miiller was followed by Alfons Stauder, who had been charged by the President of the Rrichsgrsudheitsamt with the responsibility of clarifying to the press and the legislature the ethical position of the medical profession-a task that Stauder described as of “compelling necessity” in order to maintain public confidence in the profession [46]. It is hardly feasible to summarize Stauder’s very discursive statement of some 6000 words but, admitting that some abuses were committed by individual physicians. he made the following points: It was the duty of a physician to try all possible means of alleviating or curing a patient’s disease. and this might compel him to experiment; every medical or surgical intervention was to a degree experimental: without human experimentation. medical progress would cease. Stauder then criticized certain articles referring to medical practice in a draft revision of the penal code, which he compared unfavourably to the directive of 1900 referred to above. and addressed himself to the frequent question: If physicians wish to experiment. why do they not experiment on themselves and their families’! In answer to this, he lists some auto-experiments of physicians. There follows a discussion of the borderline to be drawn between therapeutic and nontherapeutic experiments. Towards the end of his discourse, Stauder emphasizes the importance of phraseology in medical publications describing experiments on man. deprecating terms such as “feeding experiment”, ‘*experimental subject” and “materials” instead of patients, concluding with a quotation to the effect that no law. or no controlling body however strong, would prevent human experimentation, but only the physician’s professional conscience. These polemics between the press and the medical profession were not unfruitful, for on 28 February 1931 the German Minister of the Interior issued a circular prescribing “guidelines for innovative therapy and scientific experiments on man” [47]. These guidelines were recommendations not having legal force. Their provisions were that the planning and execution of new treatments must be compatible with medical ethics and the rules of the art and science of medicine. Risks should be carefully weighed against expected benefits. and the treatment should previously have been tested by animal experiments. New treatments should not be undertaken without the patient’s consent after receiving an unequivocal explanation. except in life-threatening situations where it was impossible to obtain the patient’s consent. Particular care was necessary in the case of children or young persons under 18. To take advantage of social deprivation was against medical ethics. Special caution was necessary in regard to new treatments with microorganisms. especially living pathogens. In all institutions for the care of the sick, new treatments should be carried out only by the chief physician. or with his

express authorization and under his full responsibility. Any new treatment must be fully documented on the patient’s case-sheet. together. with a statement that the consent of the patient or his guardian had been obtained after a full explanation. Published results of new treatments must show due respect for the patient and for the dictates of humanity. No experiment that could as well be done on animals should be done on man. Experiments on the dying were contrary to medical ethics. During medical rducation. every opportunity should be taken to refer to the obligations implied in undertaking new treatments or scientific experiments and in publishing their results. It will be seen from the above summary that these guidelines of slightly more than half a century ago come very close to present-day requirements. especially in regard to the stress laid on the informed consent of the subject and on the need for previous animal tests. the main element missing being any reference to independent ethical review. It is noteworthy that. as mentioned above. the Berlin Medical Board and the general press had already suggested the establishment of a machinery for the prospective review of proposals for experiments on man. THE VUREMBERG

CODE

The Nuremberg Code of 1947 comprises no signihcant advance on the 1931 German guidelines. It postulates that “voluntary consent” is a she quu HOH for experiments on human subjects, but this rather tautological term has been generally replaced by “informed consent”. It consists of ten articles, comprising provisions that are reflected in later national and international codes, with the exception that article 5 condones experiments in which there is an a priori possibility of death or disability “perhaps in those experiments where the experimental physicians also serve as subject”. No subsequent or previous bioethical code has included such a waiver. It includes no provision for independent ethical review. The code was not the outcome of an attempt to frame new principles of medical ethics. but rather a formulation, in the course of a trial for war crimes, of criteria said to be widely accepted by the medical profession. against which the acts of certain physicians carried out on prisoners might be judged. The circumstances of the Nuremberg trials had little relation to normal medical practice and research in peacetime in conditions far removed from those that prevailed in concentration camps during the war. However, they provided a stimulus to the medical profession. as represented by the World Medical Association, to assert its ethical values in the form of the Declaration of Geneva. which was amended in 1968. This was conceived as a modern version of the Hippocratic Oath and, as such. was essentially a statement of the physician’s duty to his patients, without any reference to experiments on human subjects. In 1964 the World Medical Association had promulgated the Declaration of Helsinki, which was concerned with medical experiments on human subjects. and four years later the UN General Assembly requested WHO to study the question of the protection of the human personality in the light of bio-

Human experimentation

in historical and ethical perspectives

medical developments [48]. In response, WHO prepared a report in 1974. which was subsequently issued as a publication, [49] while WMA in 1975 substantially revised and amplified the Helsinki Declaration. The WHO report was essentially a review of the ethical problems posed by recent biomedical advances, and proposed no code. The revised Helsinki Declaration is so well known that it would be superfluous to reproduce its provisions. In some of the States of the U.S.A.. notably Illinois. medical experimentation on prison volunteers had been practised for some years and during World War II inmates of one of the Illinois State penitentiaries had voluntarily submitted to experimental infections with malaria. It seems highly probably that it was the publicity given by the Nuremberg trials to experiments on prisoners that prompted the Governor of Illinois. Dwight H. Green. to appoint a committeeconsisting of three physicians. a priest, a rabbi and two members of undisclosed vocation-to review the use of prison volunteers for medical experimentation. The report of this very early ethical review committee was published in slightly condensed form on I4 February 1948 [SO]. The committee concluded that there were rules for experiments on human subjects that had “been in force by common understanding and practice”. Its main recommendations, based on such unwritten rules may be summarized as follows: must consent I. Subjects after having been informed of any hazards. 2. Experiments should be “based on the results of animal experimentation” and on “knowledge of the natural history of the disease”, and must “be such as to yield results unprocurable by other methods”. The committee accepted the principle that volunteering for an experiment was a form of good conduct meriting earlier release from prison on parole. .It warned. however, that “excessive or drastic” reduction of sentence could amount to “undue influence”. In spite of the rules “in force by common understanding and practice”. these have been transgressed in relatively recent times. as exemplified by the studies of Beecher [Sl] and Pappworth [52]. CHANGIKG

APPROACHES TO DRUG CONTROL

Considering medical practice in historical perspective. as sketched in preceding pages. it is evident that there has been a radical transformation in less than a century. and that this has been brought about by experimental. as opposed to observational, research. As Sir Edward Mellanby. then Secretary of the Medical. Research Council of the United Kingdom. wrote almost 40 years ago: “The uork of Government Departments. of medical men and of nursing staff in controlling disease can only be as good as knowledge allows it to be and this knowledge has come and can only come. by medical research” [53].

Mellanby’s formulation leaves open the question whether or not medical research necessarily implies experimentation on human subjects. However, to license new therapeutic substances for general use solely on the basis of animal experimentation would imply a mass uncoordinated experiment by individual

1437

practising physicians without any of the controls or safeguards recognized as indispensable for scientihtally planned clinical trials. Initial trials on a suitable sample of human subjects under the close control of fully qualified investigators is. in fact. the only way by which risk can be minimized. It can never reasonably be demanded that an absence of risk should be guaranteed for, as pointed out by Ehrlich in 1910, this would imply an end to all further progress. In general. such organized clinical trials of new therapeutic substances as were undertaken before World War II. as exemplified by those of the Therapeutic Trials Committee of the Medical Research Council. which was not a regulatory body. were designed to test the therapeutic claims of manufacturers for preparations already available to the medical profession, rather than as tolerance trials. A case in point was the investigation of the effect of Prontosil in puerperal fever. Results of trials that were negative were not published unless of special interest, but were communicated to the manufacturer, an example of the former being a double-blind trial of testosterone propionate in prostatic hypertrophy, which had entirely negative results. Today there is an inverse emphasis, trials of new preparations on human subjects being carried out primarily with a view to determining as far as possible whether they are safe enough to be licensed, although lack of efficacity may be an additional reason for not accepting them. With etiotropic drugs, whose target is not an organ or system of the patient. but the invading microorganism, the exclusion of the possibility that there might be undesirable organotropic effects is virtually the sole justification for a clinical trial. for if an antibiotic or other antimicrobial drug is active against a living pathogen in virro and in experimental infections in animals, it will also, in the same concentrations, be effective in naturally acquired human infections. In recent years there has been increasing emphasis, starting in the United States [54], on the need for therapeutic trials to be preceded by tolerance tests on a limited number of healthy human subjects. Such tests are undertaken without any reference to the efficacity of the drug but solely in order to exclude any harmful side-effects in human beings. such as, for example, blood dyscrasias. Such typical non-therapeutic experiments may require daily medication for long periods, with frequent clinicopathological examinations. Even non-therapeutic trials such as these cannot provide more than a reasonable probability of safety. It took over 30 years of mass medication for it to be recognized that amidopyrine could cause agranulocytosis, and even longer for the unduly prolonged intake of phenacetin to be incriminated as a cause of disorders of the blood and kidneys. Probably the synthetic drug that has been consumed in the greatest quantities. by the greatest number of peopIe,.and for the longest period, is aspirin. This drug. almost a century old, is freely available to the general public. is to be found in many millions of homes. and is not known to have produced any serious toxic effects except by deliberate massive overdosage with suicidal intent, Yet it is at least questionable whether, because

NORMAN

1438

HOWARD-JONES

of its tendency to produce minor gastric bleeding in some subjects, it would. if introduced as a new product, have passed present-day stringent requirements. Whether aspirin justifies the universal belief in its analgesic action in banal complaints is another question. It is therefore evident that even the most careful tolerance tests in human subjects. followed by scientifically controlled clinical trials. cannot entirely exclude the possibility of untoward effects in some patients. Consequently, the clinical use of such drugs remains to a limited extent experimental until extensive and prolonged usage in medical practice has shown them to be without risk. In 1975, Senator Edward M. Kennedy, in his capacity as Chairman of the U.S. Senate Subcommittee on Health, suggested that a better system was necessary “for monitoring the effects of prescription drugs after they are marketed’ [55]. As a direct result of this suggestion. a Joint Commission on Prescription Drug Use, sponsored by seven professional organizations and the (U.S.) Pharmaceutical Manufacturers Association, was established to study the problems of ‘post-marketing surveillance’ (PMS). A major recommendation of the commission was that a PMS system should be developed for the whole country. In the US. the Food and Drug Administration, and in the U.K. the Department of Health and Social Security. have already gone some way towards establishing PMS, and the World Health Organization has for some years been responsible for an international programme for monitoring adverse effects of drugs, the operational aspects of which were taken over in 1978 by a WHO Collaborating Centre in Uppsala. Sweden. It is sometimes suggested that no purpose is served by drawing a distinction between therapeutic and non-therapeutic experiments. Nevertheless, it can hardly be gainsaid that there is a fundamental difference of objectives in the two cases. The primary purpose of the therapeutic experiment is to test whether a new treatment favourably i@uences the course of a disease. The only purpose of the non-therapeutic experiment is to test whether a new treatment produces disease in a previously healthy subject. SOME SPECIAL CONTROLS

ETHICAL

PROBLEMS:

AND PLACEBOS

It has long been conceded that, given the subjective reactions not only of patients but also of physicians desirous of achieving a beneficial result, it is essential that for most therapeutic trials there should be a group receiving a different or no treatment, the attribution of subjects to one group or another being randomized. There are a few exceptions to this rule, such as the use of haemodialysis for renal failure, for which a successful result in a single case would have sufficed. It goes without saying that a control group should never consist of patients from whom a treatment of known benefit is withheld unless there is good reason for believing that a new treatment will be equally or more effective. Where there is no known treatment for a disease, it would not seem to be unethical to test the effects of a treatment by reference to a control group receiving only a placebo. Again, for symptomatic ills

for which no underlying organic lesion can be found. such as idiopathic headache. and for which the only indicator is the patients subjective impression. the use of a placebo would not seem to be objectionable. Randomized trials of treatments for serious disease. whether involving comparison of the results of a new treatment with those of a conventional treatment or of a placebo, pose the ethical problem of deciding when the trial has become conclusive-that is. whether a strong suspicion that one group has fared better than the other requires suspension of the experiment of whether the trial should be pushed to the point of statistically adequate probability. A special case is that of double-blind trials. in which neither investigator nor subject knows which subjects have received the new treatment or which have received a conventional treatment or placebo. In such a situation. the investigator who might observe apparent improvement or deterioration in some subjects by comparison with the others has no means of correlating treatment and results. It has been suggested that this dilemma might be resolved by adding to the research team an independent member who knows all the facts and can “blow the whistle” if appropriate [56]. INVESTIGATORS

AS FELLOW-SUBJECTS

There are many experiments on human subjects that carry no ethical implications, Foremost among these are auto-experiments. in which investigator and subject are one, and it has often been enunciated as a golden rule that an investigator should not undertake an experiment on a human subject that he would be unwilling to perform on himself or on members of his immediate family. This criterion loses validity when a hazardous and even life-threatening auto-experiment is combined with the same experiment on others. as in the case of the experiments of the Yellow Fever Board of the American Army. Thus. when the Board, consisting of Walter Reed, James Carroll. Aristide Agramonte y Simoni and Jesse William Lazear carried out their investigations of yellow fever in Cuba in 1900, they first considered the “moral responsibility” and “agreed that members of the Board would themselves be bitten and subject themselves to the risk that necessity compelled them to impose on others” [57]. Reed prepared quite an elaborate consent form for volunteers other than the investigators. and this guaranteed a sum of SlOO “in American gold” for participation in the experiments and a further $100 should the participant contract yellow fever, the total $200 to be paid to next of kin in the event of death. Reed, then in his late forties, had been advised. or ordered, by his military superior. not to be a subject of the experiment on account of his age! Agramonte was already immune. Carroll and Lazear were both infected with yellow fever. from which the latter died. All other subjects survived and collected their bounty. This example poses the question: Does willingness of the experimenter to expose himself to serious and possibly fatal consequences justify recourse to other human subjects’? Although the Nuremberg Code gave a guarded affirmative to this question. the answer must surely be in the negative. Even if an experimenter is ready to subject himself to serious risk. this

Human experimentation does not justify same peril. THE PLACE

his placing

OF FORMAL

another

ETHICAL

person

in historical

in the

REVlEn

Many biomedical experiments on human subjects. commonly described as ‘non-invasive’, raise no ethical problems. These include such procedures as monitoring of physiological functions by electronic or other instruments that do not encroach upon bodily integrity (unless this is done as an aid to interrogation procedures instead of for medical reasons), or exercise-tolerance tests under different atmospheric pressures. In the case of some experiments that are in a sense invasive-such as the studies of the epidemiology of scabies that were made on conscientious objectors during the last war or, after the war, the offer of free board and lodging to young honeymoon couples in return for their participation in experiments on the etiology of the commoncold--the risk is only of minor discomfort, and it would seem pedantic in the extreme to invoke formal ethical review. A distinction iS sometimes made between ethical review and scientific review, and in some current writings on the subject the impression is obtained that the forger is a necessary prerequisite to any innovative procedure on man. But there must surely be many cases in which only scientific review is meaningful. If, for example, a new chemotherapeutic agent has been found to be active against a living pathogen in vitro and in experimental infections in animals, if in the concentrations necessary for antimicrobial action it has no apparent toxic action in several animal species and if a properly constituted independent scientific review committee is satisfied by the protocols. it is difficult to see what purpose could be served by an additional ethical review procedure before proceeding to the first trials on human subjects. On the contrary. it would seem that a formal ethical review procedure is highly desirable where there is a known or suspected risk that must be weighed against the expected benefit to the individual and/or to society. It might also be considered that all nontherapeutic experiments should be subject to formal ethical review before being initiated. Here there is no question of evaluating the risk/benefit ratio. but of deciding whether it is permissible to expose healthy human subjects even to a seemingly improbable risk when it is known that there will be no benefit. In spite of Ehrlich’s affirmation of 70 years ago that there can clearlv be no difference between the therapeutic ethics of-internists and surgeons. an extensive review of the literature has not revealed any examples of formal ethical review of new surgical interventions, although these are as experimental as. and sometimes more so than. new medical treatments. An element in this discrepancy is, perhaps, that it is easier to give an intelligible explanation to the patient of the risk ‘benefit ratio of a new or hazardous surgical operation than it is to explain in terms that are meaningful to the subject the possible effects of a new drug. For example. a patient with an intracranial tumour threatened with certain blindness or other catastrophic disability if the disease were to take its natural course may well arrive at a calm and reasoned decision to risk death on the operating table-a de-

and ethical perspectives

I439

cision that no ethical review committee could take on his behalf. A special case not subject to ethical review is that of physical treatments in psychiatry, of which the most common is electroconvulsion therapy. This has neither a theoretical basis nor consistent results, and therefore each application of this treatment is largely experimental. The main argument advanced in favour of this empirical treatment is that it ‘works’. However, the history of medicine is largely composed of treatments that ‘worked’ as long as they were still used and a cynic might feel that treatments are not abandoned because they cease to work but that they cease to work because they are abandoned.

WHO

SHOULD BE THE SUBJECTS EXPERIMENTATION?

OF HUMAN

For therapeutic experiments the subjects are largely self-selected, in that they are affected by a disease or disability for which a new or putatively improved treatment is to be tested. Both patient and physician are motivated by the desire to obtain a better therapeutic result than was possible by current treatments or lack of them, although an exception to this generalization has to be made in the case of children and the mentally incompetent. Notwithstanding what has sometimes been said to the contrary, non-therapeutic experiments on healthy human subjects-r ‘Phase 1’ experiments in the terminology of the U.S. Federal Regulations-are in an entirely different category, for the incentive provided by the hope of improved health is lacking. In fact, the only positive outcome of such experiments could be the demonstration that a new treatment might be harmful to the health of the subject. This lack of an innate incentive in the form of the hope of improved health poses the question: What are acceptable catchment areas of the population from which healthy human volunteers may appropriately be recruited? The voluminous literature on this subject is conspicuous for the objections made to the of any defined social group that has been used or proposes as a source for the recruitment of volunteers. For example: Medical students might be influenced by their wish to make a favourable impression on their professors. Employees of the pharmaceutical industry may hope for greater job-security or even preferment by volunteering as subjects for experiments. Personnel of the armed forces may feel that refusal to volunteer may be regarded as verging on insubordination. Prisoners may be influenced by the desire to be installed in more comfortable quarters, to avoid violence or homosexual overtures from fellow-prisoners, by the hope of earlier parole for good conduct, or by a combination of all or some of these motivations.

The last of these groups constitutes a very special as a search of the relevant literature indicates that the use of prison volunteers has been accepted in only three countries--Israel [SS]. the Netherlands [59] and the U.S.A. In the latter country, experiments on prison volunteers attained industrial proportions,

case.

1440

NORMAN

HOWARD-JOSES

constructing some pharmaceutical manufacturers extensive laboratory facilities within the prison walls. A U.S. physician with 25 years’ experience of using prison volunteers has plausibly argued that ?he most captive of the people I have ever used have been patient-subjects”. He maintains that the patient’s ‘-great faith, confidence and trust in his physician” is “the most coercive element” for volunteers for experiments [60]. Clearly. the same special relationship of trust is not a factor in the case of health prisoner volunteers, Nevertheless, the use of prison volunteers in the United States is waning. Eight of the United States have banned it. and since January 1977 it has no longer been possible in Federal prisons. except for problems inherent in or arising from imprisonment. A suggestion that the dying are suitable volunteers for experimentation on the grounds that “an unparalleled opportunity exists for the wider study of these individuals [sic] in pharmacologic. physiologic. and other medical investigations” [61] seems unlikely to win widespread approbation. Nor does the surprising suggestion that there should be “compulsory experimentation”. justified by the curious argument that as more deaths are caused by disease than by wars, “there is no significant difference between general recruitment when a war breaks out, and the mobilization of society for the sake of fighting sickness” [62]. What is even more surprising is that this suggestion was made, not by an over-zealous medical investigator, but by a judge. Three crucial questions remain unanswered in the literature studied: First. if the various special social groups cited above are to be excluded on ethical grounds as sources of volunteers. from what other social groups is it proper to recruit volunteers? Second, if it is unethical that volunteers for non-therapeutic experimentation should be influenced by hopes of some kind of personal gain, what are the motives that would otherwise incite them to volunteer’! Third, if it is unethical to accept volunteers for experimentation in return for personal gain, is it ethical to recruit employees for high-risk occuptations that offer exceptional pecuniary rewards in compensation for the risk to life incurred. for example, by divers employed for the exploitation of North Sea oil? Society accepts without question that workers in hazardous occupations, such as coalmining, should expose themselves to the risk of sudden death or prolonged invalidism in return for a wage, and that test-pilots employed by aircraft manufacturers should take great risks for high pay: No medical experiments on healthy human subjects involve such known risks, yet this is an area in which it is widely considered unethical to offer any material compensation, except for injury that may result from the experiment. INFORMED

CONSENT

Much paper and ink have been expended on attempting to define the undefinable: ‘Informed consent’. A weakness of these attempts is that they do not take into account the semantic confusion implicit in the word ‘informed’. A person who has an ‘informed opinion’, or is ‘well-informed’. on a subject is one who has such a degree of knowledge or experience as to

enable him to make ;I considered judgement. in the light of which he can make a wise decision. On the other hand. even an Einstein L+ho has been ‘informed’ that his flight has been delabed or cancelled is a passive recipient of information that he has no means of challenging. To take another analog> : X highly educated biomedical scientist ma! ha\e :1 colour television set. his understanding of which is limited to pressing the right buttons to get his chosen programme. If the set develops a fault. he calls in :I T.V. repair-man. who explains what must be done and how much it will cost. The biomedical scientist has been ‘informed’. without necessarily having understood the explanation. and agrees to pay the bill in order to have the function of his set restored. Even the T.V. repair-man may not understand the theoretical electronics, but knows from training and expertence that a certain defect requires replacement of ;L certain pack of components. With these analogies in mind. one IS tempted to wonder what ‘informed consent’ to human experimentation. whether therapeutic or nontherapeutic. could ever mean except except to subjects already in possession of relevant medical knowledge. However. most human subjects of medical experimentation are likely to be drawn from quasiliterate. quasinumerate. hospital populations whose understanding may be in inverse relation to the completeness and accuracv of the information given to them. The British Medical Association expresses a realistic point of view in the following words [63] :

“Most patients trust their doctors and ~111consent to any proposal. Experimental procedures ;~re nearly alw;~~s too technical for patlents or non-experts to understand”. A recent inquiry showed that of 200 cancer patients invited to reply to a questionnaire on the purpose of consent forms. 159 circled “legal documents to protect the physician’s rights”. 98 “legal documents to protect the patient’s rights”. and 86 “explanations of treatments” [64]. VICARIOUS COhSENT

‘Vicarious consent’ is in a sense on contradiction in terms, for there is not consent by the subject. Moreover, the subject has normally not authorized the giver of vicarious consent to act on his behalf. Nevertheless. vicarious consent to therapeutic experiments on children or mentally incompetent subjects would seem to have no or few ethical implications if the experiment is inspired solely by the will to safeguard or improve the well-being of the subject. In the case of prophylactic experiments, such as vaccine trials. the ethical position is less clear. because prior animal experiments may be misleading or even impossible. Sharp differences of opinion exist in respect of the validity of vicarious consent in unsophisticated rural communities in developing countries in which the concept of individual rights is foreign. behavioural decisions being taken at the community level and formulated by the village headman on behalf of the entire village. In such a case. the giver of vicarious consent has the tacit authority to act on behalf of the community. It has been argued in a recent report that for the purpose of what might be described as ‘off-

Human experimentation

in historical and ethical perspectives

shore experimentation’ the process of obtaining consent must be adapted to such local customs and traditions [SS]. This point of view received authoritative-and entirely independent-support in a paper by Professor 0. 0. Ajayi of University College Hospital. Ibadan, Nigeria, on “Taboos and Clinical Research in West Africa” [66]. The author refers to the “near impossibility” of applying the concept of informed consent “among large populations of West Africa” and stresses the importance of tribal leaders in this context. However, at a CIOMS Round Table Conference held at Cascais. Portugal. late in 1978, Professor J. 0. Oluwasanmi of Nigeria vehemently repudiated such an idea, declaring it to be “completely unethical” to accept the vicarious consent of a tribal chief on behalf of the members of his tribe [67]. This is clearly a very delicate situation. the only solution to which is scientific and ethical review by the countries in which a proposed offshore experiment is to be conducted. Moreover. it is obvious that such experiments are justified only if they are motivated by the intention of finding a means of solving a health problem of importance to the population concerned. It would. however, be a mistake to assume that offshore experimentation on human subjects necessarily implies an investigation sponsored by an industrially developed country in a developing country. A recent inquiry by a British clinical pharmacologist addressed to several leading British-based pharmaceutical firms showed that there is an increasing tendency to conduct Phase II studies-i.e. the first therapeutic trials on a limited number of patients-in the United States where. in spite of stringent regulatory requirements. such studies are possible with less delays than in the United Kingdom [68]. It has been authoritatively estimated that the time required to obtain authorization for the first clinical trials in the U.K. is on the average four times that in the Federal Republic of Germany. Holland. Sweden and the U.S.A.. and that over 80”” of new drugs developed in the U.K. “went overseas for clinical testing” [69]. In the same country. the Office of Health Economics (which is financed by the pharmaceutical industry) recently estimated that the delay between synthesis of a new drug and its marketing was now about IO years, and suggested that the “one word” explanation of such delays was “thalidomide” [70]. However. “changes have recently been Introduced to facilitate the conduct of clinlcal trials of neh drugs in the United Kmgdom. These changes became necessary because early deve-

lopmental detriment

work

on new drugs

of British

departments

Industry

of clinical

RESEARCH

01

and

was going with

pharmacolog)” THE

HL’MAN

abroad

a loss

of skill

to the in our

[71]. FETUS

AND

ITS TISSUES

The use of human fetal tissue. or of the whole fetus. is indispensable for some medical research. including the culture of certain pathogenic viruses. immunological and chromosomal studies. the study of fetal development. and the preparation of certain vaccines. For example. in 1965. John Enders and Thomas Weller were awarded the Nobel Prize for growing poliomyelitis virus in cells cultured from human fetal tissues.

1441

While there are wide differences of opinion as to the circumstances in which the use of human fetal material is ethically justifiable, there is unanimity that a potentially viable fetus should in no circumstances be the subject of experimentation. The question of the period of gestation at which a fetus may have become viable has been the subject of national and international idscussion since early in the twentieth century, initially with no relation to fetal research but with a view to determining for the purpose of vital statistics at what stage a product of conception prematurely expelled from the uterus ceases to be classified as a ‘miscarriage’ (spontaneous abortion) and when it should be recorded as a ‘stillbirth’. This problem is as far from being the subject of an international consensus as it was more than half a century ago. In 1912 a committee of the Royal Statistical Society of Great Britain called for an international definition of ‘stillbirth’. proposing a period not less than 28 weeks as the criterion. In 1924 the Health Organization of the League of Nations, citing this committee’s report, issued A Preliminary Report on the Lack of Uniformity in the Dejnition of Stillbirths [72]. This was followed in 1927 by a “Committee on Definitions of Stillbirth” of the League’s Commission of Expert Statisticians. which rejected the proposal of the League’s Health Section that a period of gestation of not less than 28 weeks should be accepted as a criterion of viability and therefore of stillbirth [73]. In the following year the League reproduced a report of an Austrian committee of statisticians which recommended that a fetus should be regarded as a miscarriage when its age was less than 28 weeks, its length less than 40 cm and its weight less than 1500 g (!) c741. In 1930 the Health Organization of the League of Nations decided that for the purpose of vital statics an expelled fetus was viable if it was.the “product of gestation which had lasted at least 28 weeks” and its body-length was “at least 35 centimeters from the crown of the head to the base of the heel” [75]. There was no reference to weight. Many years later, in 1974, a WHO scientific group recommended that an expelled fetus weighing 500 g or more should be recorded as a birth, while recognizing that fetuses of 50@999 g, corresponding to periods of gestation of 22-28 weeks, had little chance of survival [76]. These limits were proposed for the purpose of international comparability of vital statistics and not in relation to fetal research. However. in 1972 a national advisory group discussing the problem of fetal viability from the point of view of fetal research. recommended in the “Peel Report” that an expelled fetus weighing 40@5OOg, equivalent to a period of gestation of 20 weeks. should be regarded as possibly. if not probably. viable [[7]], The group made the reservation that these limits might have to be revised in the light of advances in medical knowledge and the pertinence of this reservation is illustrated by; the limit of 15OOg proposed by the Austrian commtttee 45 years before. In the “Peel Report” it was proposed that whole fetuses should be used for research only if they were under 300 g in weight. The report rejected any experiments on living fetuses that were destined for abor-

1442

NORMAN

HOWARD-JONES

tion, although many such experiments have been performed. for example to determine the transplacental distribution of therapeutic substances in the fetus [78]. To accept killing a fetus while rejecting a harmless premortem pharmacological experiment upon it may appear to be straining at a gnat and swallowing a camel. but a logical, if rather implausible, justifica. tion for this stance is that a pregnant women might make a last-minute decision to refuse abortion. It was also recommended in the Peel Report that the decision as to whether a fetus might be suitable for experimental or other uses should rest with the birth (and presumably abortioh) attendants. and never with those proposing the experiment. A further recommendation was that there should be no traffic in fetuses or fetal material involving pecuniary considerations. In the United States the National Research Act of 1974 imposed on 12 July a temporary moratorium on research supported by federal funds on living human fetuses [79], but this was lifted on 29 July of the following year, when new regulations allowed federally-supported fetal research to be renewed under certain conditions [SO]. Many sectors of the ,public and some physicians find the idea of research on whole living fetuses, or even on living fetal tissue, as indefensible on moral, religious, or emotional grounds, or a combination of these. It would therefore seem that the only relevant guideline that could be proposed in this case is that when fetal research is undertaken it should be only on subjects that would, by common consent. have no hope whatsoever of continued extrauterine existence. GENERAL

COMMENTS

This review is essentially an annotated inventory of some contemporary aspects of experimentation on the human body considered against their historical background. All aspects included could be and have elsewhere been, discussed in much greater detail. Among the aspects not included here are: Behavioural research. Innovative investigational techniques, such as amniocentesis and fetoscopy which have been considered to be no longer experimental. Empirical treatments such as electroconvulsion therapy and psychosurgery, which are accepted by some but not by others. Confidentiality of medical records in relation to observational (epidemiological) research, Modalities of regulatory requirements, Of current regulatory requirements in some countries it may be thought that a very few unhappy accidents. and some entirely atypical wartime medical experiments that were neither ethical nor intended as such. have led to such a barrage of defensive restrictions as may, if further elaborated, offer a threat to medical progress. particularly in regard to industrial pharmaceutical research. It may be asserted without fear of contradiction that is as easy to cite examples of the benefits of human experimentation-which has completely transformed medical science and practice-as it is difficult to cite noxious effects. Indeed. the two outstanding

drug disasters-those of elixir of sulfanilamide and thalidomide. that were instrumental in the introduction of more stringent measures to protect the public. were due in the first case to failure to make any trials in animals and in the second to a general lack of awareness of the need for tests for teratogenicity in several animal species. The unhappy victims of both these disasters were unwittingly serving in a capacity that should have been that of laboratory animals. not as subjects of honrfide human experimentation. SUMMARY

AND

CONCLUSIONS

Apart from auto-experiments by physicians. the history of medicine before the twentieth century provides very few clear examples of human experimentation in the sense of planned investigations to test a reasoned hypothesis. Innumerable medicines of natural origin were ‘exhibited’, hopefully but without a scientifically demonstrated rationale. The writings of Claude Bernard. Oliver Wendell Holmes and Sir Thomas Clifford Allbut offer sufficient evidence of the scepticism of discerning physicians as to the efficacy of medicinal treatment in the late nineteenth century. Considerable advances in diagnosis were not matched by similar progress in therapy and such a state of the art. combined with a general lack of appreciation of elementary biostatistical principles. was hardly conducive to scientifically planned clinical trials that. in any case. would have shown most currently used medicaments to be totally ineffective. Early in the twentieth century. advances in physiology, pharmacology. and biochemistry-such as the discovery of hormones and vitamins-began to make possible a more rational approach to medical treatment. At the same time, the pharmaceutical industry was beginning to produce synthetic remedies, mostly palliatives, that were often more effective and more convenient than traditional naturally-occurring medicaments. Gradually the contributions of industry were to become preponderant in therapeutics and, ultimately, the major stimulus to the elaboration of regulations to preserve the rights of the human subjects of trials of new therapeutic agents. Some of the more important of these contributions, as also examples of the adventitious influence of the industry in the development of protective regulations. have been given, It is noteworthy that Prontosil. the first modern chemotherapeutic drug, which heralded a revolution in the treatment of common infections. was introduced on the entirely false premise that it had no direct action on the pathogen, but in some way stimulated the host’s resistance. Such a theory was in direct contradiction to Ehrlich’s formulation of the principle of chemotherapy some 30 years before. For most of the first half of the century. medical ethics did not go far beyond codes of medical etiquette, the human rights of the patient being hostage to the conscience of the individual physician and, in the last resort. litigation for injury resulting from alleged medical incompetence or unorthodox treatments. It was in the cradle of the pharmaceutical industry-Germany-that the first formulations of ethical requirements for experiments on human subjects originated in the form of a Prussian ministerial

Human

experimentation

in historical and ethical perspectives

directive laying down in 1900 the essentials of the principle of informed consent. In 1910. Paul Ehrlich re-asserted this principle in more explicit terms, comparing the duty of the experimental pharmacotherapist to that of the surgeon to inform his patient fully of the statistical risks of an operation compared to those of non-intervention. The German ministerial circular of 1931 prescribing guidelines for “innovative therapy and scientific experiments on man” was quite clearly intended to assuage widespread public concern that, inter a/b, “hospitals were working for the chemical industry”. There appears to have been no parallel for such concern, or for such governmental action, in any other country. In the United States. the Federal Food, Drug and Cosmetics Act of 1938 was a direct consequence of the public concern aroused by deaths from the ingestion of an ‘elixir’ of what was by then known to be the active principle of Prontosil-sulfanilamide. Early in the 1960s the realization that thalidomide was capable of causing gross malformations of the fetus led to a much heightened awareness of the hazards of administering to human beings entirely new synthetic substances. Seen in historical perspective, the Nuremberg Code had little relation to the development of contemporary codes for regulating human experimentation and. in particular, the scientific and ethical requirements for introducing new therapeutic or prophylactic substances into medical practice. In fact. it may be stated with some confidence that current codes or regulations for the protection of human subjects of medical research would not have been affected had the Nuremberg Trials not taken place. Imerrmrioruzd codes Both the Nuremberg Code of 1947 and the 1964 Declaration of Helsinki of the World Medical Association have been superseded by the WMA’s 1975 revision of the latter. commonly referred to as Helsinki II. which is widely regarded as the basic document in the field. In 1979 the World Health Organization (WHO) and the Council for International Organizations of Medical Sciences (CIOMS) embarked on a joint three-year project. involving multiple international consultations. for elaborating the principles of Helsinki II. The outcome of the project has recently (1982) been published as Proposed internationa/ yuidelirles $0~ hiornrdicu[ rr.warc/~ involciny hman subjects. henceforth referred to as “The WHOKYIOMS guidelines” or “the guidelines”. Brief comments are made above on some special problems of medical ethics. viz: (i) controls and placebos; (ii) investigators as fellow subjects; (iii) the place of formal ethical review: (iv) who should be the subjects of human experimentation’!: (v) informed consent: (vi) vicarious consent: and (vii) research on the human fetus and its tissues. There is no reference to controls or placebos in Helsinki II or in the WHOCIOMS guidelines. although a “General Survey” preceding the latter stresses the need in double-blind trials for an independent observer who knows the code to “oversee trends and to alert the investigator on any cause to conclude the trial prematurel!“. As regard mveitigators as fellow-subjects. while the Nuremberg Code opened the door to experiments \vith an u priori possibility of

I443

death or disability if the physicians also served as subjects, neither Helsinki II nor the guidelines offers such a loop-hole. On the question of the place of ethical review, both Helsinki II (to a very limited extent) and the guidelines seem to assume that all research involving human subjects should be subject to ethical review and make no exceptions. for example, for epidemiological research. Neither Helsinki II nor the guidelines contains any positive suggestions as to the selection of subjects for human experimentation. Helsinki II emphasizes the need for consent of legal guardians of minors and others who are legally incompetent. The WHO/ CIOMS guidelines go into considerable detail in enumerating the safeguards to be observed in the case of children, pregnant and nursing women, mentally ill and mentally defective persons, “other vulnerable social groups”, especially “junior or subordinate members of a hierarchically-structured group” and subjects in developing countries. The guidelines also contain provisions relating to community-based research, where *‘consent on a person-to-person basis may not be feasible”. None of the questions posed above under the heading “Who should be the Subjects of Human Experimentation?” is answered by the guidelines or by any national code or regulations of which I am aware. On the question of informed consent there are differences of approach by Helsinki II and the WHO; CIOMS guidelines. The former appears to accept the concept of informed consent as being an adequate safeguard, although requiring (article I, 2) that the experimental protocol “should be transmitted to a specially appointed independent committee for consideration, comment and guidance”. However. a later article (I I, 2) states: In the lo use his or lishing

treatment of the sick person. the doctor must be free a new diagnostic and [sic] therapeutic measure. if in her judgement it offers hope of saving life. reestabhealth or alleviating suffering.

If this be so. what is the function of the “independent committee”‘? Or. indeed, of “informed consent”‘? In the WHO/CIOMS guidelines, the stress is almost entirely upon independent ethical review. While articleI, 9 of Helsinki II requiring “freely given informed consent” is cited, the citation is qualified by the statement: Of itself. however. informed consent offers an imperfect safeguard to the subject and it should always be complemented by independent ethical review of research proposals. As regards vicarious consent. Helsinki provides for “informed consent” of a legal guardian in cases of legal incompetence. The guidelines also provide for vicarious “consent” of a legal guardian, but stress that in such cases “independent ethical review is imperative”. In the case of experiments involving subjects in developing countries who are not equipped to understand the implications of participation in an experimental investigation. the guidelines propose that “the decision whether or not to participate should be elicited through the intermediary of a trusted community leader”. This clause was included in the full knowledge that some spokesmen of developing coun-

HOWARD-JONES

NORMAN

1444

tries had objected to it. while proposing no feasible alternative. No mention of research on the human fetus iS made either in Helsinki II or in the WHO,CIOMS guidelines. but thi; is a subject that is doubtless more appropriately considered on its own. as has been the case in the United Kingdom with the “Peel Report” and in the United States with its very detailed Federal regulations. While there are much more rigorous and explicit national regulations governing human experimentation, especially in the United States. the WHO! CIOMS guidelmes represent by far the most complete existing international bioethical code. No international code could have the specificity of national regulations. for there must be flexibility to allow for adaptation to particular national circumstances. In conclusion. it is manifest that any new medical treatment. or preventive or diagnostic measure. whether resulting from animal experiments or otherwise, must ultimately be tested for the first time on human beings. In other words, human experimentation is the sirle quu nor] of medical progress. REFERENCES I. Bernard C. E\-pdrimwtule.

2.

3.

4.

5.

6.

Introduction d pp. 139-140.

/‘Etude

de

/u

:Mt;drcmr

Librairie Joseph Gibert. Paris [undated reprint]. Bernard explains in the preface to this classic that it is intended to be but the prelude to a more extensive two-volume work to be entitled Prrnciprs de Mtidecine E.upt++mentrt/e. By the time of his death in 1878 Bernard had not completed this xmbitious work. but left voluminous sections of it. These were located. collated, and published in 1947 by Dr Lion Delhoume under Bernard’s chosen title (see Ref. [9]). Scarborough J. Celsus on human vivisection at ptolemaic alexandria. CIio Mrdicu Il. 25-38, 1976. Abraham J. J. Lrttsom. His Life. Times. Friends und Drscmdunt.s. pp. 186-188. William Heinemann. London. 1933. Dixon C. W. S~nullpou. J. & A. Churchill. London. 1962. Dixon’s remarkable monograph includes (pp. 227-712) __ a photographic reproduction of Maitland’s published (1722) report. Dixon S. W. op. cit.. p. 261. Dixon’s monograph contains (pp. 216-295) a masterly account of the history both of vartolation and vaccination. Dieffenbach J. F. Physiologisch-chirurgische Beobnchtungen bei Cholera-Kranken. Cholera-Arch. I. 86-105, 1832.

7 Johnson wrsv.

Sir George. with

Dwection.s

History of‘ the Cholertr Contm/i)r the Treutment of’ the Diwusr.

J. i A. Churchill. London. x Willcock J. W. 711e Luws

1896.

Rrlut~q to the Mrdicul

fiwion; with un Accorrnt of the Rise und Progress ~‘urrrnts Orders. pp. 109-l IO. J. & W. T. Clarke.

Proof 1t.s

.A. L’&r dt,rt qlt’c~rn~Grl+ Stund der hue I/Choierrrfiw~e wtd &r d~shr~~i~~/~htr .St~/hst~n/~~ib’ersuche. M. Perles. Wien. 1894. 15. Pettenkofer MM.von. L’rbrr Choltwtr mat Btwk~.swhtr&en tfons

der Arzneien un yesundrn Mrn.syessumrlt. Vol. i. C. Cnoblock.

Leipzig. 1825. JGrg declared his intention of publishing further volumes every one or two years, but these never materialized.

qunq

der jiinysten

7-14.

J. ,F. Lehmann.

Miinchener

1

Cholrrarprdemie

medkinische

Miinchen.

in

Humhury.

1892 (offprint

pp. from

CVochenschr$).

Dach E. F. Selbstversuche von Aerzten mit Lebenden Krankheitserregern. Cihu-Z. pp. 147-180. 1954. Gordon H. L. Sir Jumes Yowq Simpwrr und C/~/or+

pp. lO&lO7. T. Fisher Unuin. London. 1897. I Garrison F. H. .1,1 Introduction to rhr HlstorJ r,f .)/U/Itine. 4th Editlon. pp. 565-566. W. B. Saunders. Philadelphta. 1929. 19. Forssman W. Die Sondierung des rechten Herzes. ~//JI Wscl~r. 8. 2085-2087. 1929. 20. Sonnedecker G. (Ed.) Krumers und C’rdany‘s Historic o/ Pharmacy, 3rd Edition. p. 428. J. B. Lippincott. Monform (lU//-/X70).

treal. 1963. 21. Holmes 0. W. ,Lfedicu/ Essct~~s. i&f,? /XX_‘. p. 265. Houghton Mifflin, Boston. 1883. 1 22. Ibid.. p. 186. 23. Ibid.. p. 203. 24. Allbutt T. C. Grrrl\ .Medicinr in Rome. Ii’ith Other HIStori& Essuw. p. 418. Macmillan. London. 1921. Sir Lauder. Collected Pupers U,I Circ~tiurmn 25. Brunton und Respiration. pp. VII-VIII. Mucmlllan. London. 26. 27.

28.

29.

30.

1907. Dale Sir Henry.

The natural history and chemistry of drugs. Phurmuceut. J. 152. 38. 1944. Ehriich P. Chemotherapeutische Trypanosomen-Studien. Brrl. kli,l. Wscl~r. 44. 233-236. 28&2X?. ?lO- 314. 341-344. 1907. Domagk G. Ein Beltrag zur Chemotherapie der bakteriellen Infektionen. Dr. med. Wschr. 61, 25&253. 1935. Colebrook L. and Kenny M. Treatment of human puerperal Infections and of experimental Infections in mice with Prontosil. Luncrt 1. l279- 1286, 1936. Long, P. H. and Bliss E. A. C/inicu/ L’se of Sul/unr/umide

and

Sulfupyridine

und

Allied

Compowds.

p.

8.

Macmillan. New York. 1939. 31. Domagk G. op. cit.. p. 252. 32. Trkfouel J.. Trtfouel Mme Nitti F. and Bovet D. Activiti du p-aminoph&ylsulfamlde sur les infections streptococciques experimentales de la souris et du lapin. C. r. hebd. Seunc. Sot. Biol. 120. 756758, 1935.

Long P. H. and Bliss E. A. op. cit., p. 2. 34. Chain E.. Florey H. W.. Gardner ,A. D.. Heatley N. G.. Jennings M. A., Orr-Ewing J. and Sanders A. G. Penicillin as a chemotherapeutic agent. Luncrt 2. 226-228. 1940. 35 Abraham E. P.. Chain E., Fletcher C. M.. Florey H. W.. Gardner A. D.. Heatley N. G. and Jennings M. A. Further observations on Penicillin. Lancer 2. 1977-1989. 1941. 36. Compdtrtron of Selected Acts \vrthrn the Jwlsdictron o/ the

Lon-

don. 1830. 9 Bernard C. Prrncipes de Mt;decrne Eupdrimentule. p, 18. Presses Universitaires de France. Parls. 1947. 10 lhrd.. p. 181. II Bernard C. Ref.[l], pp. 141-142. I?. Bernard C. Ibid.. pp. 263. 264. 13. JGrg J. C. G. ~Mutrriu/m x einer kiinfiiyen Hrilm~ttrllrhrr durch Vrrwchr c,hen qrwonnen und

14. Drasche

37. 3x. 39. 40. 41.

Comnuttre

on Interstute

trnd

Fwei(qn

Commrrc~,.

Washington. Government Printing Office. 1979. Editorial. J. .4m. med. Ass. 110. 1610~161 I. 1938. Ibrd. 109. 1727. 1937. Ibid. 109, I91 I. 1937. Great Britain. Ministry of Health. D+r,n~trus Crru,wd by Thulrdomide. HMSO. 1964. United States. Code o/ Frdertrl Rqlt/trtion.s. Title 2 I. Parts 3Oc499. p. 93. Government Printing Office. Washington. DC, 1979.

42. Crntrrrlb/att Prrussrn.

43. Ehrlich

der

(JW,t,,t“,l

L’,lt~rrfc/ltsoer,~u/t~~f~~

pp. I X8- 189. I90 I. P. and Hata S. Dir r~perr,nr,~t~,//u

prr der Spwrliostw

(Syphilis.

Riickfullfieber.

,,,

ChemothuruHiihnwspir-

Human

44.

45.

46. 47. 48. 49.

50.

51. 5’. 53.

experimentation

in historical

illosr, Frumhtisle), pp. 139- 141. Julius Springer. Berlin. 1910. Stauder A. Die Zulgssigkeit Lrtzlicher Versuche an gesunden und kranken Menschen. Minchener medizinische Wochenschrifl, 1931, 78, 107-112: Miiller. Friedrich. Die Zullssigkeit Prtzlicher Versuche an gesunden und kranken Menschen. Miinch. med. Wschr. 78. 104-107. 1931. Stauder F. op. cit. ReicllsHesuttdllritshlurr. 11 March 1931. No. IO, pp. 174-175. United Nationas General Assembly Resolution 2450 (XXIII). 1968. Health Aspects of Human Rights with Special Reference to Decelopments in Biology und Medicine. Geneva. World Health Organization. 1976. State of Illinois. Ethics governing the service of prisoners as subjects in medical experiments. J. Am. med.Ass. 136. 467-458. 1948. Beecher H. K. New En(lI. J. Med. 247, 1354-1360. 1966. Pappworth M. H. ffumatl Guinea Pigs. Routledge & Kegan Paul. London. 1967. Mellanby E. Medical research in wartime. Er. Med. J. 2. 352-356.

1943.

Title 21. 54. United States. Code of Federal Regulations. Parts 300499. p. 71. Government Printing Office, Washington. DC, 1979. surveil55. Culhton B. J. and Waterfall K. Post-marketing lance. Br. Med. J. 280. 1175-1176. 1980. 56. Cooper J. D. and Ley H. L. (Eds) Ethical Safeguards in Research on Hw~7uns. Vol. 5. Philosophy ond Technology qf Druy Assessment, p. 132. Smithsonian Insti-

tution, 57. Carroll ordeal 75-91.

Washington. DC, 1976. J. Quo&d by Bean W. B. Walter Reed and the of human experiments. Bull. Hist. Med. 51.

and ethical

1445

70. Editorial. Drug innovation bureaucracy. Br. Med. ./. 280, 1484-1485. 1980. 71. Griffin R. and Long S. R. New procedures affecting the conduct of clinical trials in the United Kingdom. Br. Med. J. 283. 477-479. I98 I. 72. League of Nations. Health Organization, Document C.H. 167. Geneva. 8 February 1924. (“C.H.” stands for the French name of the League’s Health Committee--“Comite d’Hygi&e”). 73. League of Nations. Health Organization. Document C.H. Expert Stat.146. Geneva, 22 December 1927. 14. League of Nations. Health Organization. Document C.H. Expert Stat./60. Geneva. 29 Februarv 1928. 75. League of Nations. Health Oiganization. Memorundum relating to the Enquiries into the Causes und Preventton qf Still-Births and Mortality during the First Year of I.,[/&. Document C.H. 820. Geneva. 31 Mav 1930.

76. WHO Scientific Group on Health Statisiics Methodology related to Perinatal Events. Report. Geneva, World Health Organization. Document ICD/PE!‘74.4. 1974. 77. Great Britain. Advisory Group on the Use of Fetuses and Fetal Material for Research. Report, London. H.M. Stationery Office. 1972. 78. Mahoney M. J. The nature and extent of research involving living human fetuses. Rrseurch on the Fetus Appendix. U.S. Department of Health. Education and Welfare. DHEW Publication No. (OS) 76-128, Washington, 1976. 79. Edwards C. C. Fetal research. Science 185. 900. 1974. 80. United States. Department of Health. Education and Welfare. Protection qf Human Suhjrcts: Feruses. Pregnant Women and in vitro Fertilixtton. U.S. Government Printing Office, Washington. DC, 1975.

1977.

of Israeli prisoners to partici58. Silken P. Preparedness pate in medical experiments. In Medico/ Experimentation (Edited by Carmi A.), pp. 59-78. Turtledove, Ramat Gan. Israel (no date). 11 June 1981. Partial trans59. Nederlandse Staatscourant, lation into English in the Int. Dig. Hlth Legisl. 32, 92-93. 1981. 60. Shubin S. Prisoner Research. The Center for the Study of Drug Development. Umversity of Rochester Medical Center. New York. 1980. M. B. The problem of experimentation on 61. Shimkin human beings: I. the research worker’s point of view. .scic,rtcr 117. 205-207. 1953. E.qwrimrntation. pp. 6-8. 62. Carmi A. (Ed.) In Medico/ Turtledove. Ramat Can. Israel (no date). 63. The Htrndhook of Medical Ethics. p. 25. British Medical Association. London. 1980. B. R.. Zupkis R. V.. Sutton-Smith K. and 64. Cassileth March V. Informed consent-whq are its goals imperfect]) realized’? ,Ye\v E~JI. J. Afed. 302. 896-900. 1980. of Medicine. c.S Partrcrputron rn Clinrcul 65. Institute Rcstwrch in Derelopiny Cowtries. pp. 4-5. National Academy of Sciences. Washington. DC. 1980. and clinical research in West 66. Ay+ii 0. 0. Taboos Africa. J. med. Etlucs 6. 61-63. 1980. N. and Bankowski Z. (Eds) Medical 67. Howard-Jones E\-p~,~ii,lrr~t~ttr~)~~ und the Protection qf Human Rights, p, 136. XII CIOMS Round Table Conference. Geneva. Council for International Organizations of Medical Sciences and Sandoz Institute for Health and SocioEconomic Studies. 1979. A. M. Assessment of new drugs: a clinical 68. Breckenridge pharmacologist’s view. Br. Med. J. 280, 1303-1305. 1980. 69. Cromie B. R’. Testing new medicines in the UK. JI R. Sock. .IIvd. 73. 3%380. 1980.

perspectives

APPENDIX THE

WHO/CIOMS

GUIDELINES

PREAMBLE

All advances in medical practice are dependent upon an understanding of relevant physiological and pathological processes and must necessarily. in the last resort. be tested for the first time on human subjects. It is in this sense that the term ‘research involving human subjects’ is used. The context in which such research is undertaken is wide and includes: Studies of a physiological. biochemical or pathological process, or of the response to a specific interventioneither physical. chemical or psychological-in healthy subjects or patients under treatment. Prospective controlled trials of diagnostic, prophylactic or therapeutic measures in larger groups of patients, with a view, to demonstrating a specific response against a background of individual biological variation. Studies in which the consequences of specific prophylactic or therapeutic measures are determined within communities. Research involving human subjects the purposes of these guidelines as:

is thus

defined

for

Any study involving human subJects and directed to the advancement of biomedical knowledge. that cannot be regarded as an element in established clinical management or public health practice and that involves either: physical

or psychological

intervention

or assessment,

or generation. storage L$J;edical information

and analysis of records containing referrable to identifiable individ-

NORMAN

1446

HOWARD-JONES

Such studies include not only planned interventions On subjects but research in which environmental factors are manipulated in a way that could place incidentally-exposed individuals at risk. The terms of reference are framed broadly. in order to embrace field studies of pathogenic organisms and toxic chemicals under investigation for medical purposes. Analogous risks are recognized to arise in research directed to other objectives, but non-medical research does not fall within the scope of this document. Research involving human subjects should be carrted out only by appropriately qualified and experienced investigators in accordance with an experimental protocol that clearly states: the aim of the research: the reasons for proposing that it should be undertaken on human subjects; the nature and degree of any known risks: the sources from which it is proposed that subjects should be recruited: and the means proposed for ensuring that their consent is adequatelv informed. The protocol should be scientifically and ethically appraised by a suitably constituted review body independent of the investigators. The guidelines proposed below will offer some countries nothing that is not already in force in one form or another. They have been framed with special reference to the requirements of developing countries and elaborated in the light of replies to a questionnaire received from 45 national health administrations and 91 medical faculties in countries in which medical research involving human subjects is as yet undertaken on a limited scale and/or in the absence of explicit national criteria for protecting such SubJectS from involuntary abuse. The replies were received from a total of 60 developing countries. human

INTERNATIONAL

DECLARATIONS

I, The first international declaration on research involving human subjects was the Nuremberg Code of 1947. whtch was a by-product of a trial of physicians for havmg performed cruel experiments on prisoners and detainees during the Second World War. The Code lays particular stress on the ‘voluntary consent’ (‘informed consent’ is now the usual term) of the subject, which is stated to be ‘absolutely essential’. 2. In 1964. the World Medical Assoctation (WMA), at its 18th World Medical Assembly. adopted the Declaration of Helsinki (‘Helsinki I’). whtch was a set of rules to guide physicians engaged in clinical research. both therapeuttc and non-therapeutic, At its 29th World Medical Assembly in 1975. the WMA revised this Declaration (‘Helsmki II’). broadentng Its scope to include “biomedical research involving human SubJCCtS”. Some Important new provisions in the revised Declaration were that experimental protocols for research involving human subjects “should be transmitted to a specially appointed independent committce for consideration. comment and guidance” (article I. 2); that such protocols “should always contain a statement of the ethical considerations involved and should indicate that the principles enunciated in the present Declaration are complied with” (article I, 12): and that reports on “experimentation not in accordance with the principles laid down in this Declaration should not be accepted for publication” (article 1. 8). 3. Both the Nuremberg Code and the original ‘Declaration of Helsmki of 1964 have been superseded by ‘Helsinki II’. the full text of which is appended. This is the baste document m Its field and has been widely accepted as such. 4. These guidelines take account of the distmction made in ‘Helsinki II’ between medical research combined wtth professtonal care (clinical research) and non-therapeuttc (non-clinlcal) biomedical research. 5. While the general prmciples laid down tn ‘Helstnki 11‘ may be rcgardcd HIS of universal validity. their modes of application in various spectal circumstances must nccess-

artly vary. The putpose of the present guldrlines IS. thrrefore. not to duplicate or amend these prtnctples. but to suggest how they may be applied in the spectal ctrcumstances of many technologically developing countrtes. In particular. the limitations of the Informed consent procedure are emphasized and issurs specific to research involving communities rather than Individual SubJects are addressed.

CONSENT

OF SLBJECTS

6. ‘Helstnki II‘ requires (article I. 9) that human SUbJKtS should not be used m medical research unless “freely given informed consent” has been elicited ‘tfter having been adequately informed of the “aims. methods. anticipated benefits and potential hazards” of the experiment and Informed that they are free to abstain or to wIthdraw from parttctpation at an) time. Of itself. however. informed consent offers an imperfect safeguard to the subject and It should always be complemented by Independent ethical review of research proposals. Moreover. there arc many mdtvtduals. including children. adults who *ire mentally 111or defective. and those who are totally unfamiliar with modern medical concepts. who are Incapable of giving adequate consent and from whom consent implies a passive and uncomprehending participation. For such groups. in parttcular. tndependent ethical review is imperative. Chi/d,_r,1 7. It is axiomatic that children should never he the subJkcts of research that might equally well be carried out on adults. However. their partictpatton is mdispensablr for research on diseases of childhood and conditions to whtch children are particularly susceptible. The consent of a parent or other legal guardian. after :I full explanation of the aims of the experiment and of posstble hazards. discomfort or inconvenience. is always necessarv. 8. To the extent that is feasible. which will vary with age. the willing cooperation of the child should he sought. after it has been frankly informed of any posstble discomfort or inconvenience. Older children may he assumed to be capable of giving informed consent. preferably also with the consent of the parent or other legal guardian. 9. Children should m no circumstances be the subjects of research holding no potential benefit for them unless with the objective of elucidating physiological or pathological conditions peculiar to infancy and childhood.

IO. Whtle no special problems of eliclting Informed consent exist in the case of pregnant and nursing mothers as such, they should in no circumstances be the subjects of non-therapeutic research that carrtes any posstbility of risk to the fetus or neonate. unless this is intended to elucidate problems of pregnancy or lactation. Therapeuttc research is permissible only with a view to improving the health of the mother without prejudtce to that of the fetus or nursIing. to enhancing the vtability of the fetus. or to aiding the nursling’s healthy development or the ability of the mother to nourish it adequately. Research directed to induced termmation of pregnancy, or undertaken in anticipation of termmation. is an issue that is dependent upon nattonal legtslation and religious and cultural precepts and therefore does not lend itself to an international recommendation.

I I. Substantially similar ethical considerations apply to the mentally ill and the mentally defcctlve ;IS to children. They should never be the subJrcts of research that might equally well he carried out in adults in full possession of their intellectual facultlrs. but the! ‘!rc‘ clearly the only

Human

experimentation

in historical

subjects available for research into the origins and treatment of mental disease or disability. 12. The agreement of the immediate family-whether spouse, parent, adult offspring, or sibling-should be sought, but is sometimes of doubtful value. especially as mentally deranged or defective patients are sometimes regarded by their families as an unwelcome burden. Where a subject has been compulsorily committed to an institution by a court order, it may be necessary to seek legal sanction before involving the subject in experimental procedures. Other

vulnerable

social groups

13. The quality

of the consent of candidate subjects who are junior or subordinate members of a hierarchicallystructured group requires careful consideration, as willingness to volunteer may be unduly influenced by the expectation, whether justified or not, of adventitious benefits. Examples of such groups are medical and nursing students, subordinate laboratory and hospital personnel. employees of the pharmaceutical industry. and members of the armed forces. Subjects

in developmy

communities

14. Rural

communities in developing countries may not be conversant with the concepts and techniques of experimental medicine. It is in these communities that diseases not endemic in developed countries exact a heavy toll of illness, incapacity and death. Research on the prophylaxis and treatment of such diseases is urgently required and can be finally carried out only within the communities at risk. 15. Where individual members of a community do not have the necessary awareness of the implications of participation in an experiment to give adequately informed consent directly to the investigators, it is desirable that the decision whether or not to participate should be elicited through the intermediary of a trusted community leader. The intermediary should make it clear that participation is entirely voluntary. and that any participant is free to abstain or withdraw at any time from the experiment. Community-based

research

16. Where

research is undertaken on a community basis-for example by experimental treatment of water supplies. by health services research or by large-scale trials of new insecticides. of new prophylactic or immumzmg agents. and of nutritional adjuvants or substitutes-individual consent on a person-to-person basis may not be feasible and the ultimate decision to undertake the research will rest with the responsible public health authority. 17. Nevertheless, all possible means should be used to inform the community concerned of the aims of the research. the advantages expected from it and any possible hazards or inconveniences. If feasible. dissenting individuals should have the option of withholding their participation. Whatever the circumstances. the ethical considerations and safeguards applied to research on individuals must be translated, m every possible respect. into the community context.

RE\‘IEH’

PROCEDURES

18. The provisions for review of research involving human subjects are influenced by political institutions, the organization of medical practice and research and the degree of autonomy accorded to medical investigators. Whatever the circumstances. however. a dual responsibility exists within society to ensure that:

411 drugs and devices under Investigation in human subjects meet adequate standards of safety. The provisions of ‘Helsinki II’ are applied in all biomedical research mvolving human SubJeCts.

and ethical

Assessment

perspectives

1447

of safety

19. Authority

to assess the safety and quality of new medicines and devices intended for use in man is most effectively vested in a multi-disciplinary advisory committee operative at the national level. Clinicians, clinical pharmacologists. pharmacologists, toxicologists, pathologists, pharmacists and statisticians have important contributions to offer to these assessments. Many countries at present lack resources to undertake independent assessments of technical data according to procedures and standards now considered mandatory in many highly developed countries. Improvement in their capability to subserve this function is dependent. in the short term, on more efficient exchange of relevant information internationally.

Ethical

review

committees

It is not possible

to draw a clear dividing line between scientific review and ethical review. for an experiment on human subjects that is scientifically unsound is ipso facto unethical, in that it may expose the subjects to risk or inconvenience to no purpose. Normally, therefore. ethical review committees consider both scientific and ethical aspects. If a review committee finds a research proposal scientifically sound, it will then consider whether any known or possible risk to the subject is justified by ‘the expected benefit and, if so, whether the proposed procedure for eliciting informed consent is satisfactory. 21. In a highly centralized administration, a national review committee may be constituted to review research protocols from both scientific and ethical standpoints. In countries where medical research is not centrally directed. protocols are more effectively and conveniently reviewed from the ethical standpoint at local or regional level. The basic responsibilities of locally operative ethical review committees are two-fold: 20.

To verify that all proposed interventions and. particularly. the administration of drugs under development, have been assessed by a competent expert body as acceptably safe to be undertaken in human subjects. To ensure that all other ethical considerations arising from a protocol are satisfactorily resolved both in principle and in practice. 22. Review committees may be created under the aegis of national or local health administrations, of national medical research councils or of other nationally-representative medical bodies. The competence of committees operating on a local basis may be confined exclusively to a specific research institution or it may extend to all biomedical research involving human subjects undertaken within a defined geographical area. 23. Local review committees act as gatherings of the investigators’ peers and should be so composed as to provide complete and adequate review of the research activities referred to them. The membership may include other health professionals, particularly nurses. as well as laymen qualified to represent community, cultural and moral values. Independence from the investigators is maintained by precluding any member with a direct interest in a proposal from participation in its assessmenL 24. The requirements of review committees should be particularly stringent in the case of proposed research involving children, pregnant and nursing women, the mentally ill or mentally defective persons. members of developing communities unfamiliar with modern clinical concepts. and any invasive non-therapeutic research. lr?formation

to be provided

bJ investigators

Whatever may be the pattern of the procedure adopted for ethical review. it should be based on a detailed protocol comprising: 25.

NORMAN

1448

HOWARD-JONES

,4 clear statement of the ObJectives having regard to the present state of knowledge and a justification for undertaking the investigation m human sublects. ,4 precise description of all proposed interventions. including intended dosages of drugs and planned duration of treatment. A statistical plan indicating the number of subjects to be recruited and the criteria for terminating the study. The criteria determining admission and withdrawal of Individual subjects. mcludkg full details of the informed consent procedure.

interest of the host country submitted with the comments the mitiatmg country. such research council or academy 29. An Important secondary sored research should be the the host country to carry independently.

COMPENSATIOrV

to require that it should be of a responsible authorIt> 01 as J health ndmmistratlon. of medicine or science. obJectIre of esternaIl> spantrainmg of health personnel of out similar research proJects

OF RESE-\RCH

FOR .ACCIDESTAL

26. There lish:

should

also be included

information

SPONSORED

RESEARCH

27. The term externally sponsored research is here used to refer to research undertaken in a host country but initiated. linanced and sometimes wholly or partly carried out by an external international or national agency with the collaboration or agreement of the appropriate authorities of the host country. 28. Such iesearcb implies two ethical imperatives: The research protocol should be submitted to ethical revie\* by the imtiatmg agency. The ethical standards applied should be no less exacting than they would be for research carried out withm the initiating country. After ethical approval by the initiating agency, the appropriate authorities of the host country should. by means of an ethical review committee or otherwise, satisfy themselves that the proposed research meets their own ethical requirements. Where externally sponsored linanccd by a pharmaceutical

SLBJECTS R\

to estab-

The safety of each proposed intervention and of any drug or device to be tested. including the results of relevant laboratory and animal research. The presumed benefits and potential risks of participation. The means proposed to elicit informed consent or. when this is not possible, satisfactory assurance that the guardian or family will be appropriately consulted and the rights and welfare of each subjtct will be adequately protected. Evidence that the investigator is appropriately qualified and experienced and commands adequate facilities for the safe and efficient conduct of the research. Provisions that will be made to protect confidentiality of data. The nature of any other ethIcal considerations involved together with an indication that the principles enunciated in ‘Helsmki II’ will be implemented. EXTERNALLY

I\Jl

research is initiated and manufacturer. it is in the

30. Reports of accidental m~ur) to sub]ects volunteering to participate in therapeutic or non-therapcutlc research and resulting in temporary or permanent disability. or even death, are excessively rare. In fact. human subjects of medlcal research are usually in exceptIonally favourable cucumstances in that they are under close and contmued observation by highly qualified investigators uho arc alert to detect the earliest signs of untohard reactlons. Such conditions are less likely to occur m routine medlcal practice. 31. However, any volunteer subjects involved in medical research who may suffer inJur> as a result of their partlclpation are entitled to such financial or other assistance .IS would compensate them fully for any temporar!, or permanent disability. In the case of death. the dependants should be eligible for appropriate material compensation. 31. Experimental subjects should not. in giving their consent to participation. be required to ualve their rights to compensation in the case of an accident: nor should the! be required to show negligence or lack of a reasonable degree of skill on the part of the Investigator. Support IS increasing for a system of insurance agamst risks. financed either by public or private funds or both. the injured party having only to show a causal relationship between the investigation and his injury. For research sponsored by pharmaceutical manufacturers. the manufacturers themselves should assume responsibility m case of accidents. This is particularly necessary in the case of externally sponsored research when the subjects are not protected bk social security measures.

CONFIDENTIALITY

OF DATA

33. Research may involve the collection and storage of data relating to individuals. uhich. if disclosed to third parties. might cause harm or distress. ConsequentI>. arrangements should be made by mvestigators to protect the confidentiality of such data. as for example by omitting information which might Iead to the identification of indlvidual subjects. by limltmg access to the data. or other appropriate means.