Research Participants and the Right to be Informed

Chapter 11
Research Participants and the Right to be Informed

Deryck Beyleveld and Roger Brownsword


Introducing Autonomy, Consent and the Law,1 Sheila McLean remarks that if the law is ‘to facilitate or protect the capacity of an autonomous person to make an autonomous choice – one that reflects his or her own values – it is necessary to develop standards that can ensure meaningful patient participation in healthcare decisions’.2 And, the development of such standards, McLean suggests, ‘is generally said to be the function of the legal doctrine of consent’.3 However, having analysed a sweep of English medical law (focusing on end-of-life questions, pregnancy, genetics and organ transplantation), McLean claims that the jurisprudence of (informed) consent is less than fully congruent with the idea that the autonomy of the patient is focal. Indeed, McLean concludes: ‘In building the rules around consent to treatment, courts have stopped short of ensuring that people are fully informed and therefore truly able to weigh their decisions and act in a self-determining manner’.4

Following McLean, we would say that, were autonomy to be regarded as focal, it would be for each individual patient, not for clinicians, to decide whether a suitable available treatment should be commenced, modified or discontinued, and which of such treatments to have. Guided by such an autonomy model, (informed) consent would translate in the following way. If a patient refuses to consent to a procedure (for example, a pregnant woman refuses to consent to a Caesarean section or a person refuses assisted nutrition and hydration5), then she should not be subjected to the procedure; conversely, if a person consents to the administration of a drug (for example, a drug that will end her life), then the drug should be administered. But, as this example shows, there may be other considerations, such as the good of other patients and society’s perceptions of the value of human life, which prevent the views of the individual patient from being the sole concern. McLean’s point is that, if consent is to serve the autonomy of a patient, it must be designed to track and reflect that particular patient’s will (however seemingly ‘irrational’ or ‘unreasonable’) rather than operate in accordance with its own rules.6

According to McLean, the autonomy of patients is compromised not only when their consent or refusal of consent is egregiously ignored but also when they make decisions that ‘are based on misunderstandings or on overt or covert denial of choice resulting from withholding of information that would have been important to that particular patient …’.7 Following up this latter claim, we might inquire how far the foreground ‘informational’ doctrines of English medical law serve the background idea that the patient’s autonomy is fundamental. This prompts McLean to review the well-known common law cases on disclosure of risk.8 However, we might extend this line of inquiry. For example, in the context of experimental procedures, we might ask whether the law requires clinicians to inform patients in a way that facilitates the latter making their own treatment decisions.9 If the law permits a clinician to withhold information about a procedure in order to steer the patient towards a treatment decision that health-care professionals would generally endorse, this suggests that a thread of paternalism remains in the law; and, by contrast, if the law permits a clinician to withhold information about a procedure in order to steer the patient towards a treatment decision that health-care professionals would generally regard as far too risky, this raises quite different concerns about the safety of patients. Either way, the autonomy of patients is compromised.

In this chapter, our interest is in the analogous question of the responsibilities of researchers to inform participants in a way that respects the autonomy of the latter. If participation is to be on an informed basis, there are many questions to be answered – for example, questions that relate to the purpose (or purposes) of the study, whether the research processes or products might be patented (and commercialised), whether and how the privacy of participants and the confidentiality of their information will be respected, whether participants will remain identifiable (and linkable to the samples and data that they provide), whether participants have proprietary rights in relation to their samples and data, and so on. Much could be written about each of these questions. However, our present interest is in the responsibilities of researchers to return autonomy-relevant findings to their (identifiable) participants – in short, the question is whether there will be (and should be) any individualised feedback of health-related findings made by the researchers. For example, if a researcher realises that a particular participant has a life-threatening aneurysm, would the law be deviating from an autonomy-centred focus if it did not require the finding to be disclosed – or, at any rate, would this be a deviation if the participant had either already consented to the return of findings or had not indicated that there should be no return of findings?

Currently, this line of inquiry is of intense interest because it takes us to one of the most complex and contested issues in the ethics and governance of biobanks. There is no single definition of biobanks, and myriad examples now exist, domestically and internationally, of collections that are of recent origin as well as historical, in some cases going back many decades. A common feature, however, is that these are collections of biosamples, personal data, scans and images, and so on. While some might be held for clinical care purposes, their wider value lies in their role as a research resource. The central question here – a question that Catherine Heeney and Michael Parker rightly single out as ‘[o]ne of the most hotly debated’ in the context of modern biobanking practice – is ‘whether there is an obligation to feedback research results to participants’.10 Suppose, for example, that researchers, who are conducting genetic analysis on biobanked materials, identify a particular mutation for breast cancer in a sample provided by an identifiable participant. Do the researchers have an obligation to inform the participant; or, to turn this round, does the participant have a right to be informed? If it is claimed that the participant does have such a right, a host of further questions need to be addressed, including questions about the scope of the right, its weight in relation to any competing rights, whether researchers owe feedback responsibilities also to third parties (such as relatives of the participant), how the information is to be conveyed to the participant, how the right might be affected by an explicit ‘no feedback’ policy at the biobank, and whether the right to be informed also implies a right that researchers actively ‘look out for’ potentially clinically significant findings.11

Responding to the headline question of whether researchers have a responsibility to feed back to individual participants those findings that might be clinically significant, we are guided by a rights-based ethic, specifically an ethic that is rooted in the seminal work of Alan Gewirth.12 According to Gewirthian ethics, in a community of rights, agents are categorically required to respect the generic conditions of agency – for present purposes, particularly the conditions of physical and psychological well-being that are essential before an agent has any prospect of successfully engaging in purposive activity (whatever the agent’s purpose, whatever the activity). The generic rights that agents directly hold against one another are both negative (against unwilled acts of interference with the generic conditions) and positive (for acts of assistance in support of their generic conditions provided that this is consistent with the agent’s will and does not threaten equally or more important generic rights of those providing assistance). It follows that our blunt – we emphasise that this is extremely blunt – response to the central question is that, where the information in question relates to an agent’s generic conditions, then researchers do have feedback responsibilities to their individual participants.

This chapter is in three principal parts. First, we sketch the context for the present debate about the informational responsibilities of researchers, particularly analysing the significance of a biobank’s ‘no feedback’ policy and the plausibility of a participant’s claim that they have a ‘reasonable expectation’ that feedback will be provided. Secondly, we outline the key features of the Gewirthian view of positive rights and its justification. Thirdly, we apply this ethic of rights to a number of illustrative scenarios where the question is whether Gewirthian-guided researchers would do the right thing by giving (or not giving) feedback to a participant; and, at the same time, we draw out the implications for anonymisation (as a strategy to protect participant privacy) if this has the consequence that researchers are thereby disabled from identifying participants who have a right to be informed.

The Context

Why is it that the question of the informational responsibilities of researchers should have become so ‘hot’? It is not that biobanks are new. However, large population-wide biobanks, established as a resource to be curated for access and use by the research community (much like a library) are new; and the possibility of undertaking (and understanding the import of) quite detailed genotyping and sequencing is assuming much greater prominence. Moreover, where the studies are of a longitudinal nature, with researchers needing periodically to recontact participants (for example, for further samples or data to be supplied, or for remeasurement, and so on), it is in the nature of the project that clinically significant findings can be returned to individual participants. If feedback can be given by biobanks and their researchers (or researchers who use the resource), the question then is whether it ought to be given.

In many biobanks, the general rule is that ‘no feedback’ will be given to participants. However, in two respects ‘no feedback’ is open to misunderstanding. The ‘no feedback’ notice notwithstanding, there might actually be some findings that are returned to participants.

First, even where the declared policy is one of ‘no feedback’, this usually refers to the position once participants’ samples and data have been ‘banked’. Prior to that point, some health-related information might be given to participants. For example, at the point of enrolment, participants might be given their blood pressure or bone density readings, or their BMI score, and so on; and, there might also be some advice about incidental observations as when a participant might be advised to check out a suspicious-looking mole with their doctor. However, once the samples and data have been collected and ‘banked’ for the use of researchers, the ‘no feedback’ policy signals that the general rule is that there will be no individual feedback arising from the findings made by researchers (general findings, of course, will be disseminated in the usual way).

The second misunderstanding is more subtle. When a biobank declares that its policy is ‘no feedback’, this might mean quite literally that there is no feedback, that there is an absolute rule against feedback. Here, ‘no feedback’ is intended to signify that participants have no right to feedback and that researchers have a duty not to inform; in no circumstances will even clinically significant results of research undertaken on the banked materials be returned to individual participants. However, ‘no feedback’ might be intended less literally, signalling something more specific, namely that the researchers do not accept any obligation to give feedback. This latter reading is designed to counter participants who assert that they have a right to be informed and that, concomitantly, the researchers have matching obligations. While such a reading is intended to shield researchers against claims for feedback made by participants, it does not preclude the giving of feedback, but, whether or not participants are informed, is exclusively for the biobanks and their researchers to decide. If the latter is the policy of the biobank, the relevant discretion might be structured in more than one way – for example, by adopting defaults of various kinds, reserving some feedback questions for case-by-case determination, and so on; and, in practice, some feedback (drawing on research on banked materials) might be given.

In the light of these remarks, consider the ‘no feedback’ policy of the Avon Longitudinal Study of Parents and Children (ALSPAC) the well-known longitudinal study of parents and their children in the Bristol area. ALSPAC declares its policy on disclosure of biomedical information to participants in the following terms:13

The policy is that information shall not, as a general rule, be disclosed to participants.

This general policy should only be set aside when it is reasonably certain that the benefits of disclosure clearly outweigh any possible risks to the participants or their families. This in turn will arise when three conditions are met:

1. That an item of data gives clear, unequivocal information of an existing or future health problem.

2. That the health problem identified is amenable to treatment of proven benefit

3. That the participant has indicated beforehand that they wish to be informed if such a problem is identified.

While the function of the ‘no feedback’ notice at ALSPAC is to scotch any idea that participants have a right to feedback, this does not rule out the possibility of informing individual participants about clinically significant findings. On this view, although participants have no right to be informed as such, they might nevertheless be informed if they have signalled this preference and provided that the researchers judge that there is a net benefit in giving individual feedback.

One of the points made in support of the ALSPAC policy (and, likewise, biobanks that have a similar policy) is that participants understood when they enrolled that there would be no feedback. In other words, the policy is supported inter alia by the following considerations:14

• Individuals have consented to participate in the study on the clear understanding that all measures are for research purposes only and not to inform decisions about their health. To emphasise this, it is often stated explicitly in the information given to participants.

• As a corollary, it is frequently repeated that the tests that participants undergo are not a check on their health, and that if participants are worried they should go to their own doctors.

• The relationship between researcher and participant differs from that between doctor and patient. Crucially the duty of care is different. The primary concern of a researcher is not to the health of a participant but to acquire information for the benefit of humankind.

There is no doubt that the ALSPAC ethics committee is seeking to do the right thing but the question is whether, from a rights-based perspective, this paternalistic policy is appropriate.

Heeney and Parker come close to answering this question. Noting that one ‘no feedback’ strategy is, in effect, to manage the expectations of participants, they say:

One route would be to make it clear to participants and health professionals at the time of consent that there will be no feedback of research results. There are a number of arguments supporting this including its potential for greater clarity about consent and about the distinction between research and clinical care and the fact that feedback assumes some sort of infrastructure in which the connection with participants is maintained to the extent that they can still be contacted and told to seek medical advice, for example.15

However, as the authors remark, this approach has fallen out of favour in those cases where research might ‘produce very clear evidence of a serious harm which might be avoided by an easily available intervention and where there exists something akin to a duty of easy rescue’.16 In other words, if our premise is that participants have a positive right to be informed (possibly akin to a positive right to be rescued where this is straightforward and not difficult for the rescuer), we are less likely to judge that biobanks do the right thing by withholding clinically significant findings. Moreover, even if a ‘no feedback’ policy has been ‘communicated’ to participants, we might wonder whether its significance has been fully appreciated; and we might judge that, regardless of the biobank’s declared policy, participants ‘reasonably expect’ to be given feedback where a biobank holds clinically significant, serious and actionable information about a particular individual.17

These observations invite two lines of inquiry. First, on what basis might we defend the premise that participants have a positive right to be informed? Secondly, how does the idea of ‘reasonable expectation’ play in this context? We respond to the first question in the next part of the chapter; and we close the present part by making some short remarks in response to the second question.

One of the striking features of much of our ethical and regulatory thinking (notably in relation to the interest in privacy) is that the recognition of a right hinges on the question of whether we judge that a person has a ‘reasonable expectation’ that his or her particular interests will be respected and protected. If a participant’s claim to have feedback hinges on whether it is based on a reasonable expectation to have feedback, then the question is: by reference to what standard or practice or to whose authority is the expectation judged to be a reasonable one?

First, the participant might invoke relevant background rules of law. There has been much discussion of whether a participant might succeed against a researcher in a tort claim for wrongful non-disclosure.18 The consensus is that English law does not clearly support such a claim; and, if the legal test turns on whether it is ‘fair, just, and reasonable’ to place researchers under a feedback responsibility, this seems merely to restate the original question of whether the claimant’s expectation is a reasonable one.

Where a biobank declares a ‘no feedback’ policy which is clearly notified to participants, if anything, this further weakens the participant’s tort claim. However, where ‘no feedback’ signals that the biobank reserves a discretion to give feedback, a participant might argue in a judicial review that the discretion has been exercised improperly. If successful, the claimant might compel the researchers or biobank to reconsider their decision; but, of course, a claim of this kind would only get off the ground if the policy set by the biobank and its administration were judicially reviewable.

Secondly, the participant might claim that the researchers had formally or informally signalled that feedback would be given. Where a biobank has a well-advertised ‘no feedback’ policy, this kind of claim would be unlikely to succeed. However, if the policy had not been signalled, then the claim would turn on whether the participant could show on the facts that an undertaking to provide feedback had been given. In principle, unless the background law prohibited researchers from giving feedback, their voluntary assumption of a responsibility to give feedback would be a strong ground for claiming a reasonable expectation of feedback.

Thirdly, the participant might rely on a general attitude that there should be some reciprocity in the relationship with researchers: participants assist researchers in various ways in return for which researchers should assist participants by giving appropriate feedback. There does seem to be evidence that at least some (and, quite possibly, many) participants sign up with the expectation that there will be reciprocation.19 However, the fact that others share one’s own expectation does not make anyone’s expectation reasonable. Possibly, the claim for reciprocity might be grounded in some other way – for example, in the way that Henry Richardson relies on the relationship of ‘entrustment’ between participants and researchers;20 but it is not enough that the de facto expectation is widely held by participants.

Fourthly, the participant might rely on the settled custom and practice at other biobanks or in a certain sector of research (or, indeed, in clinical practice as genetic analysis becomes routine). For example, it might be that researchers who work with MRI scans might consider it best practice to return incidental findings to their participants. Accordingly, where there are such practices and where the claimant participant is dealing with researchers at a biobank with no declared policy on feedback, the unstated assumption (and expectation) that there will be feedback might look perfectly reasonable. However, where ‘no feedback’ is the declared rule, contrary custom and practice notwithstanding, the argument that there is a reasonable expectation of feedback is seriously weakened.