The continuing evolution of clinical research ethics

Sloss and colleagues1 describe a qualitative investigation in which they sought the views of parents whose babies had been enrolled into a research study using ‘deferred consent’. In deferred consent, and the related process ‘waiver of consent’, the parent or legal guardian is informed about a study after enrolment and their agreement for their infants’ participation is sought retrospectively. As the procedure or treatment has already taken place, in reality this means seeking consent to use data that have already been collected, and for any continued participation. Research ethics review boards in many countries have accepted the need for deferred consent or consent waivers, because without this, studies in situations where immediate or emergency intervention is necessary would largely be precluded.

The team from the Royal Women’s Hospital in Melbourne interviewed 100 parents who had consented to a deferred request; 62 had also experienced a prospective consent process. The overwhelming majority (89) considered deferred consent acceptable. Additionally, 51 did not think prospective consent would have been preferable, 25 were unsure and 24 thought prospective consent would have been preferable. Of note, 77 parents felt they would have given the same answer if they had been approached prospectively, and only seven thought they would have declined. Parents identified their stress and anxiety following the birth of a preterm or sick baby as likely to have hindered their ability to make clear decisions. These practical and emotional perspectives underpinned their views and of importance, overshadowed any ethical considerations or concern over loss of parental rights.

The study by Sloss and colleagues is a useful addition to the growing body of evidence justifying the continued evolution of research ethics. Ethical considerations, like so many societal perspectives, are not fixed and immutable, but alter over time.2 It is worth reflecting on some of these changes. Twentieth-century considerations were primarily focused on protecting participants from the dangers of research and the rights of individuals to provide autonomous consent. These were spearheaded by the revelation of atrocities in the name of ‘medical research’ that took place during the Second World War, the Nuremberg trials and the 1964 World Medical Association Declaration of Helsinki. The Helsinki Declaration has undergone nine amendments, most recently in 2013, providing objective evidence of the evolution of research ethics.

Over the decades, the concept of ‘fully informed’ consent has been codified in ever longer patient information sheets. These place responsibility on the investigator for the task of providing ‘full’ information and induce anxiety in the patient or parent at having to read and comprehend complex detail. In contrast, several studies including that of Sloss and colleagues identify discussion with a trusted healthcare professional of greater importance to parents than autonomous decision-making.3 4 Whether in clinical practice or research, let alone research during a time of great stress and anxiety, it is questionable whether patients or parents can be ‘fully informed’. Perhaps a better focus is recognition that good clinical research, like good clinical care, requires honesty and trust. The studies that used deferred consent described by Sloss and colleagues all involved practices in established use. Thus, provision of verbal explanation with opportunity to opt out (rather than opt in), and short written information, would have been a reasonable way of reducing the anxiety of decision-making at a stressful time. The urgency of the situation would guide whether this was prospective or retrospective. Opt-out and short information sheets have been accepted by the UK Health Research Authority as an acceptable approach in defined circumstances.5 Sloss et al conclude with emphasising that ability to make a considered decision under less stressful circumstances is crucial. A challenge for research regulators and researchers is to ensure that this duty of care towards parents is recognised and incorporated into research design.

Where else might we wish to see further evolution of research ethics? What other changes might promote rigorous clinical research, given this is essential to advance care, while continuing to uphold the cardinal principles of research ethics? Box 1 provides four areas for consideration. I propose that for studies that aim to resolve uncertainties around treatments or practices already in wide use but where the evidence of efficacy or safety is inadequate or their relative (comparative) effectiveness is uncertain, it is ethical to consider allocation by randomisation as a standard of care or default, in preference to allocation by clinician bias, increasingly represented by the imposition on all patients of a non-evidenced consensus or expert opinion-based guideline. Patients risk harm from exposure to poorly evidenced treatments and practices. Conversely, randomisation provides each patient with the fairest chance of receiving the (unknown) best treatment, and thus meets the cardinal ethical principles of justice, beneficence and non-maleficence.

The Helsinki Declaration even in its most current version refers to infants and children only in relation to their need for special protection, and not as groups that also have a right to benefit from research. This paternalistic view, though well intentioned, would be better replaced by consideration of the balance of benefits to risks from research participation and non-participation. Where a study involves treatments in accepted use, there are no added risks from participation. Indeed, there may be inclusion benefit regardless of allocation arm, because care will be delivered along a closely monitored pathway. Acceptance of randomisation in defined situations as standard of care and in the patient’s best interests would also reduce the risk of rejecting research participation and any associated benefit, because of an unjustified perception of risk.

Poorly designed studies and those with inadequate power to detect clinically important effects will not deliver meaningful knowledge and therefore in addition to being wasteful, there are legitimate questions to be asked about whether they are ethical. This Helsinki Declaration quotes the words of the Declaration of Geneva, ‘The health of my patient will be my first consideration,’ and those of the International Code of Medical Ethics, ‘A physician shall act in the patient’s best interest when providing medical care.’ Few neonatal Cochrane Library meta-analyses have a sufficient information size to permit reliable conclusions to be drawn and too many included studies are low quality. Hence, patient care continues to be compromised by dangerous uncertainties and the high risk of drawing misleading conclusions from meta-analyses of small and/or poor-quality trials.

I hope to see more done by the clinical professions and their organisations to inform the public at large, and all healthcare professionals, about the purpose and power of research. I look forward to a time when parents, patients and their physicians prefer randomised allocation to resolve uncertainties in widely used treatments and practices. I suggest it is time for a further revision of the Declaration of Helsinki. I would welcome replacing the word ‘subject’ with ‘participant’ to underscore the partnership that should exist between researcher and patient. I would like to see a recommendation that research ethics committees consider the benefits as well as the risks of participating and of not participating in any given study and require investigators to articulate these clearly and honestly. I would like a better balance of protection and paternalism to ensure investigators and sponsors deliver equity so that all patients regardless of race, age, sex, or whether they are pregnant, or breast feeding are able to benefit from research participation.6 Above all, I would like to see good clinical research considered as integral to patient well-being as good clinical care.