Article Text

Download PDFPDF

Lest we forget… research ethics in children: perhaps onerous, yet absolutely necessary
  1. Joe Brierley1,2,
  2. Vic Larcher2
  1. 1Paediatric and Neonatal intensive Care Unit, Great Ormond Street Hospital for Children, London, UK
  2. 2Clinical Ethics Service, Great Ormond Street Hospital for Children, London, UK
  1. Correspondence to Dr Joe Brierley, Paediatric and Neonatal Intensive Care Unit, Great Ormond Street Hospital for Children, London WC1N 3JH, UK; brierj{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Research ethics and children

Medical research involving children is essential for advancing child health and well-being.1

Research involving children and their families still attracts controversy. In a recent article published in this journal, there has been concern voiced at the layers of bureaucracy that attended the ethical review of a seemingly straightforward but important piece of research on the follow-up of children with cardiac disease.2 To some extent, the frustrations encountered by the researchers should now have been lessened by the implementation of the findings of the Warner report3 and the establishment of a National Research Ethics Service.

Nevertheless, the ethics of medical research involving children remains a matter of balancing protection against access.4 Some have suggested that a recent domination of protection has contributed to a situation in which children are ‘therapeutic orphans,’5 this term reflecting the lack of significant evidence for many aspects of current paediatric therapeutics. For example, in emergency childhood situations, the paucity of evidence for treatment regimes, due to the lack of randomised control trials in emergency medicine and paediatric intensive care compared with other areas of medicine, means that the very sickest children are subjected to evolved, anecdotal, medical practice. Similar non-evidence-based practice in adult medicine allowed lignocaine prophylaxis postmyocardial infarction and routine episiotomies until recent well-designed randomised controlled trials exposed their dangers!

However, the recent welcome address of this imbalance with strategies such as increased funding for paediatric research, legislation to encourage—and occasionally enforce—pharmaceutical companies to test drugs in children, and the setting up of UK MCRN have led some to question whether the pendulum may have swung too far away from protection of the weak and vulnerable.4 This is especially troubling as increasing globalisation affords greater research access to less empowered populations.6

A history of research ethics

Thomas Percival, a physician from Warrington composed arguably the first ethical research code (1803). He suggested experimental techniques should be: scrupulously and conscientiously governed by sound reason' and ‘undertaken after consulting other physicians.7

Despite his novel use of the term ‘medical ethics,’ Percival made no suggestion of consent. This seems to have been first considered by the US civil war surgeon William Beaumont, who, after studying a soldier with a traumatic gastro-cutaneous fistula, enabling ready access to gastric secretions, suggested that the voluntary consent of any subject is required, and any experimentation should be discontinued if the subject became distressed or dissatisfied.8

This approach hardly became standard practice, the view of Bernard, the founder of modern physiology and experimental medicine, prevailed. He characteristically observed: It is our duty and our right to perform an experiment on a man whenever it can save his life, cure him or gain him some personal benefit. … So, among the experiments that might be tried on man, those that can only harm are forbidden. Those that are innocent are permissible, and those that may do good are obligatory.9

Such paternalism was clearly consistent with contemporary medical practice and social mores, and it is important to consider subsequent ethically questionable practices in this context, notwithstanding the extent to which any universal ethical principles might arguably transcend such justifications. In many ways the development of research ethics has been reactionary, with medical ethical norms often lagging behind social concepts of appropriate medical behaviour. Indeed, only recently, revelations from the Alder Hey scandal10 showed research using organs retained from deceased infants without parental consent. Clearly, the far past has no monopoly on ethically dubious research.

Failure to obtain consent or to properly inform subjects is not a new phenomenon. It is interesting to compare that aspect of the Alder Hey controversy with the furore surrounding research a century earlier by Albert Neisser, the discoverer of gonococcus. Neisser, Professor of dermatology and venereology in Breslau, Germany, had hoped to induce immunity against syphilis by injecting syphilitic blood serum into eight females, three of which were children, the remainder sex workers. These subjects were neither informed of nor consented to the process. When some contracted syphilis, Neisser concluded his ‘vaccination’ had failed, as the subjects had contracted syphilis due to continuing prostitution.11

The press became aware of this work, and the ensuing scandal led to widespread public debate and Neisser's prosecution, which triggered a Prussian Ministry of Education decree requiring informed consent from all research subjects, while prohibiting experimentation on minors and adults who were not fully competent.12

Furthermore, ethics training for medical students became compulsory, and several ethical codes were generated, constituting, somewhat ironically, the earliest and clearest pronouncements on consent in human research. A 1931 German Reich Minister of Interior directive forbade ‘innovative therapy’ unless: subject/legal representative unambiguously consented to procedure in light of relevant information given in advance.13

Of course, it was the war crimes tribunals following World War II that led to the Nuremberg Code and the start of modern research ethics era.14 Experiments were performed, often with fatal results, on vulnerable people in the name of science and included infection with malaria, smallpox and cholera, sexual sterilisation and exposure to extreme cold; however, despite clear evidence, many of those complicit progressed to glittering medical careers. Heinrich Gross, an Austrian doctor accused of killing of at least 800 children as part of the Third Reich's euthanasia programme, kept body parts from his ‘patients,’ including whole heads, and boasted of a ‘unique collection’ used until the late 1980s for research that secured him an international reputation in neurology.15 His trial in 2000, based on fresh evidence, was postponed indefinitely when he was ruled medically unfit to stand, and he died a free man in 2005.16

Gross practised in the Spiegelgrund clinic, Vienna, where patients were considered to be ‘unworthy of life’ because of stutters, learning difficulties and harelips, and occasionally because their parents demonstrated resistance to the regime.

While never convicted by a court, Gross's uncontested comment is chillingly clear: No child's death is in vain. I'm doing research that will be of benefit to children in the future.17

One of the novel recommendations that still remain from the Nuremberg Code today is that all research proposals must be submitted to an independent research committee. However, the Nuremberg code did not carry the force of law and was certainly not always followed. Furthermore, the Nuremberg code's most influential clause, that: ‘…voluntary consent of the human subject is absolutely essential’ effectively forbade research with children, the unconscious or those mentally incapable of consent, and for many years it was felt that research with children was unethical, if not illegal. In 1964, the Helsinki Declaration of the World Medical Associations ‘superseded’ the Nuremburg Code and included provision for proxy consent, thus permitting research on those who could not consent for themselves. This code has been revised five times, most recently in 2000, and is still the overarching document governing research with human subjects today. Contemporary UK research ethics applications still require that researchers promise to abide by the Declaration of Helsinki.18

Contrary to common perception, the Nuremberg and Helsinki declarations had little immediate effect on research regulation. Indeed, in the UK, the prevailing view was that Nuremberg merely highlighted the effects of state interference in medicine, rather than the consequences of unregulated research. The medical hierarchy considered that researchers could be relied upon to act well.19

In reality, it was revelations of controversial research practices involving children from both sides of the Atlantic that shocked mainstream physicians and led to current research regulation and the requirement for formal ethical review by a Research Ethics Committee (REC). The first of these revelations came from Henry Beecher, a US anaesthesiologist, who in 1966 described over 20 studies with dubious ethical provenance that had been performed by researchers with excellent reputations and published in major medical journals.20

Two particularly notorious studies are worth specific mention. At Willowbrook school, hepatitis viruses were deliberately injected into mentally disabled children to review the ‘natural history’ of the disease, while in another study, children, including ‘mental defectives and juvenile delinquents,’ were given increasing doses of Triacetyloleandomycin, an antibiotic to treat acne, to determine whether hepatic dysfunction—detected by blood tests and, in some cases, multiple liver biopsies—occurred.21 The Willowbrook researchers subsequently asserted that parental consent had created ethical justification, but this was firmly rebutted by Paul Ramsey, an influential ethicist, who argued that conditional admission to Willowbrook, which was otherwise refusing admissions, was an unethical inducement, and, furthermore, suggestions to parents that an immunisation programme was in place were untrue.22

In the UK, Maurice Pappworth, in his 1967 book Human Guinea Pigs: Experimentation on Man23 devoted an entire chapter to unethical research performed on children. Pappworth highlighted a number of troubling experiments from the UK, including researchers in Sheffield who had injected intravenous ammonium chloride into infants with pyloric stenosis to reverse acidosis, using children with hydrocephalus without metabolic derangement, as they were ‘readily available’ as controls.24 In Birmingham, paediatricians challenged a girl with phenylketonuria with a high phenylalanine diet without informing her mother and titrated her observed behavioural deterioration against serum assays. An unconsented control maintained both behaviour and normal levels of phenylalanine despite receiving the same diet.25

In an influential statement, Pappworth suggested: No physician is justified in placing science or the public welfare first and his obligation to the individual, who is his patient or subject, second. No doctor, however great his capacity or original his ideas, has the right to choose martyrs for science or for the general good.23

Perhaps the most notorious of all research controversies best highlights the challenge of separating research shortcomings from contemporary medical practice viewed many years later. In the American Tuskegee Syphilis Experiments (1932–1972),26 curative antibiotics were withheld from 399 out of 600 black men with syphilis who were told that they were being treated for ‘bad blood—meaning syphilis, anaemia or fatigue—in order to study the natural history of that disease.’ Compensations such as free medical examinations, meals and burial insurance seem inadequate, even when considered in the context of 1930s postdepression Deep South, with syphilis endemic and a paucity of government resources. It is, however, a damning indictment of black people's position in American society, that such work was freely published annually without furore until 1972—well after the effectiveness and relative cost of penicillin were known. The ethical abhorrence of this study seems only compounded by the fact that, once again, children were victims, it later being determined that while treatment was withheld, 40 women were infected by their spouses and gave birth to at least 19 babies with congenital syphilis.

While so-called Cold War experiments with atomic testing and subsequent radiation exposure have led to unwitting suffering by the children of indigenous peoples as well as of service personal, the radioactive cereal given to young male ‘science club members’ institutionalised at Fernald School in Massachusetts to determine the unknown effects of radionuclides in humans in the 1950s deserves special mention, as the parents, as well as the children, were simply told they were being fed a diet rich in iron.27

Although some of the researchers in the studies above may have had career progression, reputation and power as goals, it is likely that the majority of those involved—then as now– had at least some desire to further mankind's knowledge of the human condition. However, even without the advantage of hindsight, it is clear that the balance between protecting human subjects and advancing medical knowledge was somehow lost.

Although these cases are illuminating, they perhaps obscure the ethical arguments about the use of children as research subjects. Arguably, the key ethical deliberation was played out between two American theologians, Paul Ramsey and Richard McCormick,28 in the lead-up to the US National Commission for Protection of Human Subjects and the Belmont Report.29

The fundamental issue in paediatric research is, of course, the ability of others to give valid consent on behalf of someone who is unable to do so. This is especially pertinent in research that does not offer benefit to the child in a narrow medical sense, so called non-therapeutic research.22 Ramsey suggested that it is the competent informed consent of a research participant, together with the good aims of the researcher that allows them to become joint adventurers in a common cause, and thereby makes research, that does not confer narrow clinical benefit, ethical.22 Ramsey argued that as children cannot consent, and no one—not even a parent—can give informed consent for another person, then ‘non-therapeutic’ paediatric research is de facto unethical. This follows because not even a parent has the moral authority to submit a child to experiments from which that child cannot possibly benefit. Crucially, to submit a child to non-therapeutic research treats the child as merely a means (to obtain scientific knowledge), rather than as end in itself, and in so doing breaches one of the central tenets of Kantian ethics.

The contemporary protagonist to Ramsey, McCormick, also a theologian, was identifiable with a more consequentialist outlook. He argued that without research, children overall suffer far greater harm and that non-therapeutic research on children is ethically permissible as parental consent is valid, in so far as it is the best reasonable presumption of a child's wishes.30 Furthermore, McCormick suggested that the use of children as research subjects in non-therapeutic research was consistent with the Kantian principle above, as children should not be considered as separate individualistic entities, but rather as an end that is actually a social being.31

Essentially, the ethical arguments about research in children centre on whether the stronger duty is to protect them from actual or imputed harms or whether the wider beneficial consequences of performing scientifically valid research should prevail. Perhaps the side of this debate one falls on identifies one's own general core philosophical leanings, rather than views on this topic per se. However, to ignore the consequences of discouraging paediatric research seems foolhardy, and Ramsey's solution—namely, that researchers should ‘sin bravely,’ and not to deny the moral force of the imperative they violate—is therefore unsatisfactory.22

The harm, of unverified therapies, remains a factor in paediatric medicine to this day, with a critically ill child with septic shock and multiple organ failure benefiting from few therapies submitted to verification in appropriately designed trials.

Harris and Holm suggest we should not ‘presume moral turpitude in children,’32 arguing convincingly—as McCormick did—that we should consider children as being of a social group which they would automatically want to assist, by participating in research to benefit children as a whole. However, we agree with Ross in sounding a note of caution, as the pendulum swings towards ever greater access to children for medical researchers. Therefore, the need to balance overpermissiveness against unreasonable restriction, both of which may produce short- and long-term harm to vulnerable subjects, remains an active debate, as many REC discussions will attest.

Research subject to ethical codes

All human research studies must now be ‘vetted’ by an independent REC, which comprises both expert and lay membership, the latter being those not from the research community. There are two UK specialist paediatric committees, the Royal Liverpool Children's Hospital committee at Alder Hey in Liverpool and the Institute of Child Health/Great Ormond Street committee in London.

The task of each REC is to review work and assess whether it complies with ethical guidance into research with humans. There are a number of specific guidelines that can assist both researchers and any REC reviewing paediatric projects, with the Royal College of Paediatrics and Child Health33 and the Medical Research Council1 the most influential, though there has recently been criticism that the former might be excessively permissive in the research it permits as ethical.34 Crucial in this debate is the question as to whether the distinction between therapeutic and non-therapeutic research is sustainable, but this is beyond the scope of the current article, although in the examples cited, it seems clear that individuals did not benefit. There is also extensive guidance on research ethics, and the application process, on the website of the National Research Ethics Service (, part of the National Patient Safety Agency. However, the Declaration of Helsinki remains the key document with which all human research studies must comply16

Most journals insist that work has been successfully reviewed by a REC before publication; indeed even identifiable case reports usually require the explicit consent of subjects.

While the research ethics community strives to facilitate its processes, for instance with fast-track applications,35 REC applications together with internal university processes for own account research and research registration may seem an almost endless and bewildering series of hurdles. In this article, we have not sought to undermine the efforts of those who seek to carry out research in children that attains appropriate ethical and scientific standards, nor to vilify them. Rather we have sought, to achieve understanding ‘lest we forget’ of the need for ethical review and attempted to place this in some social context. Hopefully, with increasing coordination and electronic communication, a system is developing which minimises the burden of applications, and achieves greater mutual understanding between researchers and the research ethics community. However, the primary function of such bodies remains the protection of human subjects, whatever their age or status.



  • Funding None.

  • Competing interests JB is Chair of the Great Ormond St/Institute of Child Health Research Ethics committee, and also part of the Paediatric Intensive Care research group at GOSH. VL is a member of the RCPCH Ethics committee.

  • Provenance and peer review Not commissioned; externally peer reviewed.