The duty to rescue

Harris claims that one of the most powerful obligations that individuals have is to do no harm. According to him: ‘Failing to prevent harm is as effective a way of ensuring that harm occurs, and hence as morally reprehensible, as doing harm directly’.7 From this he derives the duty to rescue in the context of research: Because medical research is necessary to relieve the suffering that exists due to illness, we have an obligation to further it.

Brassington does not agree that failing to prevent harm is as blameworthy as causing harm directly, stating: ‘Naturally, from time to time there might be something morally problematic about leaving the world unaltered. But it is not qualitatively the same as positively making a deleterious alteration to the world. One person does not make another worse off simply by not making him better off’.9 In the absence of a pre-existing agreement this seems reasonable; I do not make you worse off by not buying you a car, if I have not promised to do so.

Furthermore, according to Brassington, if there is a duty to rescue, it is not clear that it gives rise to an obligation to promote medical research. Because ‘…even if engaging in research is a form of rescue, it is not the only form of rescue, and we might still have a moral reason to pursue the other forms ahead of research’.8 Since time and money may be better spent on rescuing people in other ways, it is not necessarily blameable not to take part in research, which implies that it cannot be a duty. Similarly, if a duty to pursue rescue through medical research exists, it is not clear that it must be discharged by participating in research projects, because there are other ways to support research than by becoming a research subject, for instance by donating money or taking a job as a researcher. So, although it is possible that a duty to rescue may give rise to a specific obligation to participate in research under given circumstances (eg, during a public health emergency), it seems not to be a justification of a general duty to participate.

Two other more specific duties might initially look promising, but are questionable. First, it can be argued that a duty to rescue gives rise to an obligation to engage in the kind of research that does not cost individuals anything or require active participation, and thereby does not hinder them from attempting alternative ways of rescue simultaneously. On this line of reasoning, becoming a researcher or donating money to research would not relieve people of their putative duty to further medicine for instance by accepting that tissue samples and data are used in research. However, even this duty is debatable due to the difference between doing harm and doing good. While the duty not to do harm is a perfect duty that cannot be overridden by self-determination, the duty to do good is imperfect, that is, it is largely up to the duty holder's discretion how it is executed.

Second, it can be claimed that the duty to participate does not apply to all, but only people that are in a similar situation as the person in need of rescue (just as individuals commonly assist each other at seas or when climbing mountains). De Castro has accordingly argued that the ‘commonality (in terms of their vulnerabilities and the requirements for dealing with such vulnerabilities) between the prospective subjects and the study's potential beneficiaries provides a basis for generating a duty to serve as research subject’.10 Although such a limited duty may seem more plausible, it does not give rise to the general duty that Harris wants to defend.

Free-riding and fairness

According to Harris we all benefit from the social practice of medical research, and because we accept this benefit we have a duty based in justice to contribute to its existence. Not doing so is free-riding, which is unfair.

Brassington8 ,9 does not agree with Harris that those who benefit from research without participating in research studies act as free-riders. He contends that they do pay for the benefits they receive from scientific research through taxes, insurances or upon receipt. He goes further asserting that even if they were free-riders, free-riding is not morally worrisome, stating that: ‘people who do not support research, irrespective of whether or not they are free-riders, do not make the world a worse place’.9 He also claims that ‘…appeals to fairness will only stand a chance of getting anywhere when we do actually want and expect to benefit from the fruit of that research. Should we be indifferent about it, it would be hard to sustain an argument that we are acting unfairly by not supporting it…’9 But all of these assertions are contestable.

First, as Harris notes in his reply, medical research cannot be carried out using money alone. Other contributions are necessary, such as being a participant or donating samples.7 Furthermore, the idea that individuals have a right to decide how they want to contribute to research is problematic since it imposes a risk of bias that threatens the validity of the results (because individuals that agree to participate may differ from those who do not).11

Second, people who do not support research do make the world a worse place than it would have been had they contributed their fair share to the common quest of improving healthcare. Schaefer et al2 use an example that illustrates this point: people living in a neighbourhood that want to invest in a speed bump in order to make the streets safer. The authors claim that the residents that do not contribute although they benefit from the increased safety (that they desire) act as free-riders, because if they had contributed, each individual would have had to pay less. But they assert that this argument is not valid in the context of medical research because the burden is not alleviated on those who participate, if more people contribute. However, there is another perspective to consider: more or better research can be carried out if everybody participates, just as more or better speed bumps can be put in place if all neighbours contribute. Thus, even if those who refrain from participating do not increase the cost that others must pay, they decrease the benefit that could be obtained if all individuals that wished to benefit did their share. In the case of speed bumps there may be only a certain number or quality necessary, but the need for medical advances is arguably unlimited.

Third, although it seems reasonable that appeals to fairness require an expectation of benefiting from the research, Brassington's definition of ‘that research’ as a specific research programme is debatable. The fruit of the research may as well be drawn from the existence of an effective research enterprise as such.11 If samples and data from all are used on the same terms in order to improve healthcare, it does not seem relevant whose material has given rise to which results. That is, it is not necessary to specify from what study we expect to benefit and in what way; we can simply hope to benefit as much as possible from the research that is carried out using everybody's samples. His assertion that unexpected benefits from a research programme do not generate obligations fails for the same reason.

To conclude, Harris is arguably right in claiming that considerations of fairness underlie the duty to participate in research. However, for Harris the requirement to be fair is coupled to a virtually unlimited moral demand to make the world a better place by preventing harm. In the following we will attempt to strengthen the argument by explicitly pointing out how a duty to be fair could be generated, without appealing to far-reaching foundations of moral sacrifice. We do so by situating medical research in the framework of a hypothetical contract grounded on self-interest.

A contract based in self-interest

The contract view starts with the recognition that there are certain goods that individuals cannot attain on their own, but that require cooperation. If the good is sufficiently important, it is rational even for purely self-interested individuals to agree to mutually constrain their freedom in order to acquire it. This is arguably true of effective healthcare because health is fundamentally valuable to all. Since for many diseases treatments and cures do not exist, and we do not know in advance what kind of medical care we (and those we care about) will come to need, we all have an interest in medical advances. Such advances can only be achieved through research. Research requires cooperation, and for us to cooperate as efficiently as possible, unnecessary hurdles must be avoided.

Using the example of biobank research, requiring consent to store and use tissue samples and data consumes time and costs money. It also imposes a risk of invalidating results because when participation is not universal, registries and biobanks are affected by both those who participate and those who do not, resulting in incomplete information and data-bank bias.12 So if individuals maintain a right to decide what can be done with their samples and data (which may seem to lie in their best interest), research is impeded and advances in healthcare may be delayed or even blocked, which is arguably not in anyone's interest. This is the classic problem of collective action: The choice that leads to the best outcome for an individual when considered in isolation gives rise to an outcome that is suboptimal when the decisions of all are combined. The result is suboptimal both compared with the best possible outcome for the individual and, more importantly in this context, compared with other outcomes that are achievable through cooperation. Therefore, it is rational to accept a contract imposing a duty to participate, when the associated risks and costs are diminutive. It is worth noting that individuals also have a more direct interest in participating in biobank research because future treatments for those whose genes are not represented in research are likely to be compromised as personalised medicine moves forward.13

However, there must be limits to the duty to participate, because the benefit of accepting the contract must outweigh the costs and risks involved for each individual ex ante. This is why it can apply only to certain research. Of course in reality it is not clear that the benefit will always outweigh the cost for every person. Some individuals may suffer harm, and these need obviously not be the same individuals who would have been harmed anyway had the research not been carried out. But the risk (understood as both the degree and likelihood) of harm imposed by agreeing upon the contract must stand in relation to the anticipated benefit for each individual.

This is clearly illustrated by the case of vaccinations: If many people are vaccinated some will inevitably suffer from side-effects that may be serious, and it is not likely that these will be the same individuals that would have suffered harm from the disease had there been no vaccinations. Although a vaccination programme may be successful from the perspective of the population, for a specific individual the consequences can be devastating. But even for that individual accepting the immunisation programme was rational, if the risks imposed by there not being an efficient programme outweighed the risks imposed by the immunisation.

Thus, it is not the actual outcome for an individual that determines whether or not accepting the contract is rational (because it cannot be known in advance), but the alternatives that are available at the time it is agreed upon. That is, the risks involved in accepting the contract must be weighed against the risks imposed by its non-existence.

It is important to note that the purpose of the contract is not to maximise the public good or utility in general, but to optimise the outcome for each individual, as compared with a situation in which all pursue their own best interests independently. So, this is not a case of prioritising the good of society (or of a majority) over that of individuals (or a minority), because the contract serves to benefit each individual. Consequently, it does not collide with the widely endorsed principle that the interest of the individual should always prevail over that of society in medical research.14

Considerations of fairness

It now seems clear why fairness is relevant to consider. Again, using the example of biobank research, agreeing to routinely use samples for research increases the benefit that can be achieved. However, if an individual was to oppose contributing in practice, it is unlikely that it would hinder the specific research that he or she would later come to need, because there are many other samples to use and many different kinds of research are carried out. It is only in the unlikely event that the individual's sample would have been the specific sample necessary for the success of the specific research project that would have lead to the specific advances necessary for him or her that not contributing would be an inferior decision from the point of view of pure self-interest, when considered in isolation. The problem is obvious: Accepting the contract grants us our second best outcome, but by deviating when others adhere we may obtain our first.

This applies to many societal agreements. Consider the case of pollution: If we agree that fresh air and clean water are important goods that we wish to preserve, and in order to do so accept a contract aiming to hinder us from polluting, the individual who knowingly cheats by not adhering (because he or she realises that his or her negative contribution will have no significant impact on the situation as a whole) is choosing the action that leads to his or her own best outcome, while at the same time relying on others to forego theirs.

But this line of reasoning is open to everybody, and if all engage in such behaviour, the benefit of increased medical knowledge, or an unpolluted environment, cannot be attained. We end up worse, approaching the situation where all pursue their own self-interest independently again. Relying on others to do their part while personally reaping the benefits from not adhering is free-riding, which is unfair, because it amounts to not paying for what you receive and would agree to pay for, if cheating was not possible. Thus, applying a social contract view offers a plausible explanation of why fairness is relevant to consider in the context of justifying a duty to participate in medical research. But a duty to adhere can be defended even if considerations of fairness are not admitted.

The duty to adhere

Traditional contract theories can be divided into two classes, the first presupposing a prior obligation to honour agreements, and thus to adhere to acceptable contracts (eg, exemplified by the social contract of John Locke). The second type of theory makes the contract in itself a ground for complying (eg, the theory of Hobbes); that is, it attempts to explain the duty to adhere as such.

From the latter perspective, David Gauthier has introduced the concept of ‘constrained maximizers’, distinguishing between persons who are disposed to straightforwardly fulfilling their interests in the particular choices they make, and constrained maximisers who are disposed to comply with mutually advantageous moral constraints, provided they expect the same compliance from others.15 He argues that the net advantage that constrained maximisers gain from cooperating is larger than that which persons who exploit others may expect, and thus it is rational to ‘be disposed to constrain maximising behaviour by internalising moral principles to govern one's choices’.15

The contract that we propose is based in self-interest, that is, it is rational for individuals to accept. A moral duty to adhere can be justified either by a pre-existing obligation to honour one's agreements (a widely endorsed moral principle) or by referring to the rationality of doing so. It need not, but can, be enforced by legal instruments, for example, a parliamentary decision to establish quality registries and biobanks within healthcare.

Potential objections

An obvious objection to the suggested view is that there is in fact no contract, and that individuals cannot be bound by a contract that they have not consented to. Although it is of course hypothetical, there are many other instances in which we are morally (and often also legally) bound to social contracts that we have not actively consented to; as citizens we are for instance obliged to pay taxes, although we have not signed a contract agreeing to do so. Another potential objection is that we are not (all) self-interested benefit maximising creatures, but actually care about others, and that the proposed contract therefore rests on false premises. However, we have intentionally grounded the duty to participate on the weakest possible assumption, that is, individuals care only about themselves, because allowing for considerations about the well-being of others will only make the argument for a duty stronger. It might also be objected that there are alternative contracts that would be more rational for individuals to accept, for example, one that imposes a duty to participate on all others but oneself. However, the other individuals would not accept such a contract (because it would be inferior from the point of view of their self-interest) so it would fail and thus not be rational at all.

There are also general objections against a duty to participate in medical research, such as the fact that all individuals do not have (equal) access to healthcare or that research does not guarantee advances in clinical care. Although these arguments are clearly relevant in the context of a duty to participate in research that imposes significant costs or risks (that must be outweighed by corresponding benefits), they pose less threat to a duty to participate in low-risk research. It is also worth stressing that people who have limited access to healthcare benefit from medical advances too, as is exemplified by the increased use of drugs against HIV in developing countries.