After a trial is over: the ethical issues
Soon after the completion of a clinical trial – once the initial
excitement has died down, and the results have been presented – investigators
and their sponsors are understandably keen to move on to tackle new problems.
But it is now generally accepted that their ethical responsibilities for the
population involved in the previous trial do not cease on its
The extent of these responsibilities and the ways in which they can be met are a matter of current debate. There is consensus that trial subjects should not be worse off than they would have been had they not participated in the study. More contentious is the issue of how much they, or the community from which they come, should subsequently be ‘rewarded’ for having taken part in the trial.
What the guidelines say
It is now generally accepted that the ethical responsibilities of research investigators, and their sponsors, for participants in a clinical trial do not cease on its completion. The precise nature of these responsibilities is under debate, however, and international ethical guidelines give only limited advice on what should be done. For example, the WHO guidelines on good clinical practice (GCP) state that the investigators have a responsibility to trial subjects once the trial is over, although these are not defined in any detail. [ref. 1]
The World Medical Association’s Declaration of Helsinki, states that "at the conclusion of the study, every patient entered into the study should be assured of access to the best proven prophylactic, diagnostic and therapeutic methods identified by the study", but it does not provide advice on responsibilities to control subjects or to the community at large. [ref. 2] And it does not specify what 'assured of access' means in practical terms, or who has the responsibility to assure it.
The guidelines of the Council for International Organisations of Medical Sciences (CIOMS) puts the onus on trial sponsors and investigators, stating that: "Before undertaking research in a population or community with limited resources, the sponsor and the investigator must make every effort to ensure that … any intervention or product developed, or knowledge generated, will be made reasonably available for the benefit of that population or community". In the commentary accompanying the guideline it is noted that the issue of 'reasonable availability' is complex and needs to be determined on a case-by-case basis. [ref. 3]
Similarly, the European Commission’s European Group on Ethics in Science and New Technologies suggests "a guarantee of a supply of a [proven beneficial] new drug at an affordable price for the community". [ref. 4] In contrast, last year a group of developing-country public health officials and researchers emphasised in an article in the journal Science that post-trial benefits should be conceived of much more broadly than simply 'reasonable availability' of the experimental drug or vaccine. [ref. 5]
The UK Nuffield Council on Bioethics and US National Bioethics Advisory Commission have reviewed in detail the ethical issues arising after completion of a trial. Their reports stress the importance of early consideration and discussion of post-trial issues by all the parties concerned. [refs. 6, 7]
Trial outcomes and disseminating the results
A key issue in deciding what needs to be done after a trial is over is the outcome of that study. Ethical issues are simplified if the intervention being tested is not effective (or no better than existing measures), or if it is so effective that there is a clear scientific argument to justify its introduction as rapidly as possible into routine use.
In many cases, however, the outcome of a trial is much less clear-cut. For example, a vaccine that gives only 50 per cent protection might be useful in some circumstances but not in others. It has been pointed out that an HIV or malaria vaccine with 25 per cent efficacy might be cost-effective, but could undermine public confidence in immunisation programmes when some of the vaccinated individuals develop infections.
Additionally, trial investigators, vaccine or drug manufacturers, and international and local public health officials may differ in their perceptions as to how successful a trial has been. It is therefore essential that possible post-trial implementation strategies, covering a range of outcomes, be discussed as widely and early as possible before a trial is started.
It should be noted that companies are under increased pressure to test vaccines at earlier stages of development than has previously been the case, in order to expedite their introduction into developing countries. This increases the risk of a company deciding not to seek licensure – for commercial or other reasons – of a specific vaccine formulation that has been shown to be effective during the course of a trial. It can be argued that trials of prototype vaccines that may not ultimately be registered, even if they are shown to be effective, may be acceptable (proof of principle trials) provided that risks are fully understood by local health officials, sponsors and the study community at the onset of the trial.
Disseminating the results of a trial to the scientific community through international meetings and journal publication is an essential part of clinical research. But it is equally important that results – even when they are disappointing – are presented to the trial participants and to members of the community in which the study was conducted, for example through personal contact and community meetings. Failure to do so is likely to discourage the community from participating in future studies.
Investigators should also ensure that the results are thoroughly discussed with the health authorities responsible for the community in which the trial was performed. [ref. 8] As health ministries in many developing countries do not have well-developed information systems, this may require running workshops with the relevant health officials, as well as those from neighbouring communities. In such situations a balance must be struck between conveying the success of the trial and putting undue pressure on local authorities to divert limited local resources away from other, perhaps more important, interventions.
Sustaining improved healthcare in the community
In resource-poor countries, undertaking a trial to the standards required for good clinical practice often requires strengthening the existing health service. For example, it may be necessary to provide additional staff or equipment to a hospital or to enhance delivery services such as those required for vaccination. Such resources may also allow local staff to take on additional responsibilities that were not previously possible. There is a danger that once a trial is over – and the extra resources provided for it are withdrawn – a study community could be worse off than it was before the trial began.
It is often argued that investigators and their sponsors have an ethical responsibility to take measures to ensure that this does not happen. [ref. 3] On a practical level, additional staff provided by the trial should be withdrawn gradually with appropriate notice, and support can be sought for the maintenance of equipment, at least for an initial period. When a trial has led to overall improvements in healthcare, investigators can assist local health authorities in finding ways to sustain some of these improvements, for example by reallocating staff responsibilities or by encouraging changes in working procedures.
Problems of this kind are reduced – although not eliminated – when long-term investment is made in a community where a series of trials are conducted in sequence rather than as ‘one off’ studies. Such an approach also allows long-term surveillance of study populations at a moderate cost.
Provision of treatment
In many cases there is a clear argument for making treatments that have proved successful in trials available to the local population – including those who originally participated in the research. For example, there is little ethical debate that if a new treatment for an acute, short-term illness is shown to be safe and more effective than the current treatment, and if it is affordable, then it should be made available by the public health service (although this is a relatively unusual scenario).
It is also generally accepted that if a trial shows that a new treatment for a chronic condition such as diabetes, hypertension or HIV is substantially more effective and/or less toxic than the standard existing treatment, it should be provided to trial participants for as long as necessary. If local health authorities recommend provision of the new drug as the first choice of treatment then this objective will be achieved. However, if this is not the case then there is an ethical obligation on investigators and sponsors to find ways of making the study drug available to patients in the trial for as long as they need it.
As the number of patients involved in therapeutic trials is usually relatively small, the pharmaceutical company that makes the new drug may be prepared to take on this responsibility (although it can still involve substantial financial commitments – consider the provision of anti-retroviral drugs to several hundred patients for their remaining lifespan).
If health authorities decide not to introduce the new drug into the public health service, despite it being more effective than standard care, the ethical issue arises of whether researchers and their sponsors have any responsibilities to the community beyond the research participants. A related issue is whether it is ethical to start a trial of a new drug in the absence of a firm commitment from the health authorities that the new treatment will be adopted if an agreed level of efficacy is shown (with the danger that important trials may not be undertaken as a result).
One approach (advocated in both Nuffield and NBAC guidelines) [refs. 6, 7] is that while such prior commitment from health authorities is highly desirable, it should not be an absolute requirement provided that the issue has been fully discussed with all parties – including the trial participants, as part of the consent process – before the trial starts.
In their Science article, Emmanuel and colleagues [ref. 5] suggest that a narrow focus on provision of the new treatment for an indefinite period should not be an automatic ethical requirement for undertaking a trial. Instead, they argue that this is only one among a number of potential ’fair benefits’ – such as improvements to the local health infrastructure or provision of training – that the community might directly or indirectly receive from participation in the study.
Following up the long-term effects of vaccines
Trials of vaccines or other preventive measures often raise more complex post-trial issues than those considered above. Three groups of subjects need to be considered – those who received the intervention, those in the control group, and members of the community (district, region or country) in which the trial was undertaken.
The benefits of following up the health of subjects that have received an intervention in a clinical trial are two-fold. First, it enables assessment of the long-term health impacts of an intervention to be carried out before it is registered for widespread use (an aspect that has received increasing interest in many countries). Secondly, investigators may feel they have an ethical obligation to maintain long-term follow-up of subjects in the intervention group on safety grounds. The possible need for this is shown by the fact that in 1993 a delayed increase in mortality was reported in children in Senegal and Guinea Bissau who had received a high-strength measles vaccine. [ref. 9]
Such long-term follow-up is particularly important in trials of interventions against infections whose clinical manifestations are influenced by age, such as malaria. In highly endemic areas of Africa, cerebral malaria – which has a high mortality rate – is seen most frequently in older children. [ref. 10] Thus, there is a theoretical danger that an intervention, such as a malaria vaccine, that provided protection against the disease during the first year of life could, by impairing the development of natural immunity, lead to an increase in mortality in subsequent years.
The duration of follow-up required to establish the safety of a successful intervention will vary from situation to situation, but may be several years. It is important to note that if a trial is extended beyond the initial period proposed, consent must be obtained from subjects in the trial for their participation in continued surveillance.
What happens to control subjects?
If an intervention tested in a placebo-controlled trial is found to be effective, careful consideration is needed as to whether or not the intervention should subsequently be offered to the controls. If the control subjects remain at high risk (for example from HIV or malaria), it can be argued that there is an ethical duty to offer them the intervention, with costs being met by the trial’s sponsors.
However, if the control subjects are no longer at significant risk – for example, because they have moved out of the high-risk age group or because an outbreak has subsided – then it is not logical to provide them with the intervention, even though they might expect it. Such circumstances raise difficult ethical questions. For example, is it right to provide a marginally beneficial intervention to a control child who has moved out of the at-risk age group, while denying the same intervention to his/her younger sibling who is at high risk but who was born too late to participate in the trial?
Another problem is that provision of an effective intervention to the control group makes it impossible to determine the duration of the protective effect observed in the intervention group. Thus, there may be a conflict between the scientific desirability of maintaining the status of the control group and the best interests of those in this group. This was the case for a trial of insecticide-treated bednets in Burkino Faso, reported in 1997, where providing control participants with treated bednets at the end of an initial period of observation made evaluation of the long-term impact of this intervention difficult. [ref. 11]
It is important that before a study begins, participants know whether or not 'effective' interventions will be offered to those in the control group. This information should be provided in the consent form, which should also define effectiveness. In this way, control subjects in a trial of the malaria vaccine SPf66 were offered vaccination, even though the vaccine showed only a marginal efficacy (31 per cent), because this had been promised at the start of the trial. [ref. 12]
Offering prevention strategies to the wider community: who pays?
As well as deciding what happens to those who have received an intervention during a clinical trial and those in the control group, it is important to consider members of the community (i.e. district, region or country) in which the trial was undertaken. Of course, if a new preventive strategy is shown to be effective and the Ministry of Health decides to incorporate it into the national public health service, then the community where the trial was conducted – as well as the rest of the population – will benefit.
However, the financial capacity to introduce a new intervention, even after a ‘successful‘ trial, may be lacking in many developing countries. In a worst-case scenario, the intervention will not even be made available to the study community once the trial is completed – even though the results of the trial might be used to support the introduction of the intervention in other countries. International discussions of equity in research ethics often incorporate discussions of ‘distributive justice’ (namely, the fair distribution of the benefits and burdens of research). Circumstances in which research has been conducted within one population, only for the results of it to solely benefit those in other regions or countries, have therefore attracted criticism.
One option is for a manufacturer to offer to supply the new intervention to the Ministry of Health at a discounted ‘affordable’ price (i.e. one similar to that of other vaccines purchased by the Ministry). If the Ministry is not willing to make any contribution towards these costs, it could be argued that the intervention is not deemed important enough for that country, leaving investigators and sponsors free of further responsibilities. If, however, the Ministry agrees to pay a proportion of the cost – indicating its political commitment – but still finds the ‘affordable’ price to be beyond its means, manufacturers and trial sponsors will be under increased pressure to contribute to the costs.
In some circumstances it may be appropriate to seek external support in order to cover these costs, for example from the Global Alliance for Vaccination and Immunisation (GAVI) or the Global Fund for HIV, Malaria and Tuberculosis. It is imperative that such organisations are consulted prior to the start of the study so that they can determine if the trial design and clinical endpoints address their own public health goals.
Determining responsibilities to the 'community'
In many ways, the ethical responsibility of manufacturers partly depends upon the nature of the trial that has been undertaken. If the trial has significantly helped the manufacturer to sell its vaccine in rich countries – where a substantial profit can be made – its responsibility to the trial community is high, or the trial may be viewed as exploitative. This factor persuaded a combination of public and private sources to provide hepatitis B vaccine free, or heavily discounted, to the Gambian Ministry of Health for several years after the completion of a successful trial there.
On the other hand, if a new vaccine is directed at a predominantly developing-country market and is likely to make little or no money for the manufacturer, or if the trial is not expected to enhance its marketability in industrialised countries, it is more difficult to argue that the company has an ethical responsibility to provide the vaccine for free or below-cost after the trial is over. Nonetheless, the principle of "affordable price" still applies.
Another difficult area is determining the extent of the 'community' that should benefit from the introduction of a new intervention, particularly if some or all of the costs are being borne by external organisations. For example, a Ministry of Health may question whether it is fair, or politically acceptable, to introduce an intervention such as vaccination in just one district or region when there are other areas equally at risk. In such circumstances, the Ministry may insist on national implementation or no implementation at all.
If a company agrees to provide a vaccine on completion of a trial, how should it decide whether that vaccine should be provided regionally or nationally and for how long? If the expectation is that national provision is required, there is a danger that investigators and their sponsors will only undertake trials in small countries, such as The Gambia, where post-trial commitments will be relatively modest. The ethical issues involved in making these difficult decisions are discussed by Emmanuel et al., who argue that only those regions or citizens involved in a study should share in the benefits. [ref. 5]
In the paragraphs above we have focused on delivery of a new vaccine, as vaccination is the area in which many of the questions considered above have been most extensively debated. However, the principles considered apply equally to other public health interventions in poor countries, such as nutritional supplementation or the provision of drugs at the community level for the prevention of infectious disease.
Conclusions: the importance of post-trial considerations
Participation in a clinical trial usually requires substantial commitments from the volunteers who join the trial, and from their community. Unfortunately, many consent forms only discuss the possible risks and benefits during the course of a trial, and do not address the issue of what will happen after the trial is over. It is imperative that consideration is given to what participants are entitled to after the trial is over, as well as during the course of the study.
This is particularly important when one considers that subjects enrolling in a trial usually do so in anticipation of receiving some benefit during the trial itself, not because of a promise of possible benefits after the trial is over. This important facet of informed consent is rarely taken into account in discussions of the post-trial ethical obligations of the trial investigators or sponsors.
Finally, although the nature of participants' rights will vary from trial to trial – and may be influenced by its outcome – there is an ethical obligation on all groups involved to take the issue of post-trial considerations seriously, as soon as a study is initiated. The time to start discussions on the issues considered in this paper is not after the trial is over, but before it has begun.
Brian Greenwood is at the Department of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine. William P. Hausdorff is with Wyeth Vaccines Research, United States. Both authors have been personally involved in clinical trials conducted in developing countries for many years. Their views do not necessarily represent those of Wyeth Vaccines.
 World Health Organisation. Guidelines for good clinical practice on pharmaceutical products. 1995
 World Medical Association. Declaration of Helsinki: ethical principles for medical research involving human subjects. Paragraph 30. 2000.
 Council for International Organisations of Medical Science (CIOMS). International ethical guidelines for biomedical research involving human subjects. Guideline 10. 2002.
 Opinion of the European Group on Ethics in Science and New Technologies to the European Commission. Ethical aspects of clinical research in developing countries. February 2003.
 Emmanuel E et al. Fair benefits for research in developing countries. Science (2002) 298:2133-4.
 Nuffield Council on Bioethics. The ethics of research related to healthcare in developing countries. 2002.
 National Bioethics Advisory Commission. Ethical and policy issues in international research: clinical trials in developing countries. 2001.
 Council for International Organisations of Medical Science (CIOMS). International guidelines for ethical review of epidemiological studies. Principle 13. 1991
 Aaby P et al. Long-term survival after Edmonston-Zagreb measles vaccination: increased female mortality. J Pediatr (1993) 122:904-908.
 Snow RW et al. Severe childhood malaria in two areas of markedly different falciparum transmission in East Africa. Acta Tropica (1994) 57:289-300.
 Habluetzel A et al. Do insecticide treated curtains reduce all-cause child mortality in Burkina Faso? Trop Med Int Hlth (1997) 2:855-862.
 Alonso PL et al. Randomised trial of the efficacy of SPf66 vaccine against Plasmodium falciparum malaria in children in southern Tanzania. The Lancet (1994) 344:1175-81.