Contesting the science/ethics distinction in the review of clinical research

Abstract

Recent policy in relation to clinical research proposals in the UK has distinguished between two types of review: scientific and ethical. This distinction has been formally enshrined in the recent changes to research ethics committee (REC) structure and operating procedures, introduced as the UK response to the EU Directive on clinical trials. Recent reviews and recommendations have confirmed the place of the distinction and the separate review processes. However, serious reservations can be mounted about the science/ethics distinction and the policy of separate review that has been built upon it. We argue here that, first, the science/ethics distinction is incoherent, and, second, that RECs should not only be permitted to consider a study's science, but that they have anobligation do so.

Recent policy in relation to clinical research proposals in the UK has distinguished between two types of review: scientific and ethical. Scientific review broadly includes issues to do with methodology, consideration of the context of previous relevant research, and the appropriateness or need for conducting the particular study under consideration. Ethical review is concerned with the impact of the research upon the potential participants and society at large in terms of the balance between potential risks and benefits, and the adequacy of the information to be provided to participants. This distinction has been formally enshrined in the recent changes to research ethics committee (REC) structure and operating procedures, introduced as the UK response to the European Union Directive on clinical trials.¹,² This approach was confirmed and endorsed in the recent ministerial review of RECs,³,⁴ and continues to be upheld in the recent consultation document from the Central Office for Research Ethics Committees.⁵ Whilst this latest document suggests a slight modification to the process of science/ethics review, the principle of separation remains. The planned new ‘triage' system for proposals is supposed to pick up any inadequacies in the science before they go to RECs for ethics review. It remains the case that in general RECs are expected to accept the science as presented.

However, serious reservations can be mounted about the science/ethics distinction and the policy of separate review that has been built upon it. We argue here that, first, the science/ethics distinction is incoherent, and, second, that RECs should not only be permitted to consider a study's science, but have an obligation do so.

1. The Coherence of the Science/Ethics Distinction

We begin our discussion of the science/ethics distinction by arguing that decisions about methodological issues cannot be clearly separated from their ethical implications. If this is true, then we have good grounds for questioning the coherence of the distinction. An example will help to make the point.

A recent study published in the British Medical Journal looked at the relative merits of shampoos and mechanical treatments for head lice.⁶ This study suggested that the particular mechanical device tested was a more effective treatment for head lice than medicated shampoos. However, this research was criticised in an accompanying editorial in relation to the randomisation technique and the fact that only one dose of shampoo was used (contrary to relevant recommendations).⁷ If these criticisms have merit (and whether they do is beyond our expertise), the result of allowing this research to proceed is that its conclusions may be scientifically invalid.

However, poor science has serious ethical consequences. First, the participants will have undergone potential harm by being exposed to the intervention (in the above case, any risks from the treatments, such as side effects and lack of efficacy due to the reduced dose – perhaps even the emergence of resistance), as well as a potential lack of benefit by being deprived ‘standard' care, and the general inconvenience of inclusion in the study.

Second, society has been deprived of potentially useful information from which other current and future patients are now unable to benefit – indeed, they too may suffer the harms of being treated with inadequate therapies as a result of the methodological flaws of the study.

Third, anything that brings clinical research into disrepute is damaging. Poorly conducted studies will not provide the best possible results, and support for poor research could potentially hinder future studies by making it harder to recruit to potentially beneficial research because of the damage to patients' goodwill.

Finally, research is expensive. Researchers are therefore obligated to maximise the benefit from their studies and minimise their use of scarce resources. Poorly conducted research wastes money, deprives other worthier studies of funds, and may lead to future societal costs by encouraging the use of inadequate treatments in preference to more effective ones.

2. Consideration of the Science in RECs' Ethical Review

Imagine a proposal for conducting the head lice study⁶ coming to a REC, with an accompanying scientific peer review arguing in support of the study going ahead. However, members of the REC want to question the value of the study on the scientific grounds suggested in the editorial.⁷ Could the REC reject the study? The implication of the science/ethics distinction and the current policy built upon it is that the REC could not, as such a judgment would involve the REC's conducting its own ‘science' review, and this is held to be inappropriate.³,⁵

However, we suggest that in such a situation the review of the science and ethics cannot be kept in two separate spheres. A REC, sent a study subject to such methodological flaws, should be able to reject it. Such a rejection could be justified partly on the basis of the science, whatever an expert review may have said, but it will also be based upon the ethical considerations outlined above. If a study cannot achieve its stated aims, or those aims could be better achieved in another way, it is unethical for it to go ahead. These “science‐based” methodological problems are ethically significant, because it is clear that changes to the methodology would have a positive impact upon the ethical acceptability of the study.

This, in turn, raises the issue of the quality of the scientific reviews as they are presently conducted. These range from in‐house approval by colleagues to in‐depth review by grant awarding bodies. Reviews' quality may suffer for a number of different reasons. The researchers themselves often select reviewers, raising the possibility of mutual approval of proposed research protocols. It might also be that both researchers and reviewers share a particular approach to research that makes them blind to possible problems. It is far from clear to us that the ‘triage' system, and the involvement of ‘Research Ethics Advisors', will address this issue.⁵ Why think any individual can possibly have the breadth of scientific expertise necessary? In general, RECs have a wider range of expertise and a more detached viewpoint than the group conducting the scientific review. We are not suggesting that RECs should dispense with scientific review or that they should conduct such reviews themselves on an ad hoc basis. Clearly it is imperative to get expert scientific opinion where it is appropriate. However, it does not follow from this that RECs must never question the ‘science'. We suggest that it should be, at the very least, permissible for RECs to question the researcher about methodological issues, and, if necessary, refuse to approve the project until satisfactory answers are obtained.

Indeed, we can go one step further and argue that it is not just permissible for a REC to reject a study if it is based on poor science, but that given the REC's role in weighing the potential harms and benefits of the research under review, it is obligated to do so. This argument places a strong emphasis on the REC to take responsibility for the consequences of the research it approves.⁸ For example, one important part of the remit of the REC is to protect patients from the harm that may arise from poor research. Such harm might result directly from a particular study or, more generally, if poor quality research is permitted to proceed, as we have discussed above. Any REC faced by a poor scientific basis for a study is, in our view, obligated to reject it, for these reasons.

The argument so far has reached two conclusions. First, that the science/ethics distinction is incoherent. If this is correct, then instructing RECs not to comment upon the ‘science' of a research proposal is inappropriate. Second, we have suggested that any thorough ethical review is required to consider the study's science as part of the ethical review. At least part of the justification of this requirement is the moral obligation placed upon the REC to consider all of the ethical issues relevant to a study.

3. Possible Objections to this View

Is there anything else that could offer a justification for supporting separate science/ethics reviews?

One possibility is a concern that because of RECs' amateur status, they will be unable to understand or make a legitimate judgment about the science, and so they should never be asked to do so. However, why should this be a worry? In our experience RECs are often able to point out proposals' methodological flaws not spotted by ‘experts' in the field. The ideal REC will have a broad range of experience and expertise (including both clinical and lay).

It might also be suggested that a REC just needs to ensure that potential research participants are fully informed about any problems with the science. Whilst such an approach would allow the REC to focus on the ethics (as attempting to ensure informed consent is an ethical issue) we believe that such an approach would be an odd one. Surely it would abrogate the REC's responsibility in relation to protecting potential participants from harm?⁸

Another possible argument in favour of the science/ethics distinction and separate review is the idea that science can be done independently of ethics. The model here is that the real work is done by working out what is scientifically appropriate and then trying to get as much of that as possible through the REC. According to this view, ethical review is seen as an annoying interference with the purity of the research. However, in our view ethics should not be seen as something that comes into play once the science is worked out. Ethics should be not only a part of good science, but its driver – for the primary justification for much medical research is the ethical obligation to provide the most effective treatments. In addition, much of the justification for scientific review is, in the end, ethical. For example, scientific review might suggest the study has been carried out before, or that there is inappropriate basic science to justify the proposed study at this point. These ‘science' reasons are however clearly based upon ethical claims.

One other possible justification for separate review is that it is an obligation imposed by relevant international guidelines governing research ethics. However, this is not the case. It should be noted that the science/ethics distinction is not required by the EU Directive, the Helsinki Declaration or the Council for International Organizations of Medical Sciences (CIOMS).²,⁹,¹⁰ On the contrary, all, rightly, see ethical and scientific issues in clinical research as being deeply interdependent. For example, the Helsinki Declaration suggests that a REC should review an experimental protocol (B.13). Likewise, the EU Directive suggests that the scientific issues should be considered by the ethics committee (article 6). Whilst the CIOMS (guideline 2) suggests that a science review should take place, it does not suggest it is separate from an ethics review: quite the contrary. In fact, the origins of the science/ethics distinction lie much closer to home, in the Governance Arrangements for NHS Research Ethics Committees (GAfREC). This document suggests that science is reviewed by experts (2.3; 2.3.2), although this need not be done by separate committees. This is then taken further by the Warner review in a number of places (e.g. p.7; p.13; p.15). This suggests that, if anything, the current procedures, in enforcing a science/ethics distinction, fall foul of international regulations, rather than being supported by them.

Conclusion

We have argued that the distinction between science and ethics is incoherent, and that therefore the policy of two distinct forms of review built upon it is unjustified. On the contrary it is perfectly appropriate (indeed in some situations it will be morally required) for RECs to reject studies on the grounds of poor science. The science is relevant to the ethical judgment to be made, and the attempt to maintain a distinction between the science and the ethics review is not just fallacious but also potentially unethical.