Research ethics broadly writ: beyond REB review.

• Author: Anderson, James A.
• Position: Research ethics board - Canada

"... no aspect of research, from formulation of initial hypothesis down to final clinical acceptance, is devoid of ethical significance." (1) In Canada, as well as elsewhere, governments, sponsors, institutions, and researchers tend to view research ethics almost entirely through the lens of research ethics board (REB) review. According to these stakeholders, research ethics is REB review. In this paper, we aim to show that there is much more to research ethics; health research is a multi-stage process and each stage has unique ethical implications that may or may not fall under the mandate of the REB. We also aim to show that our current governance system fails to address those ethical issues lying outside of the mandate of the REB. Finally, we consider two possible explanations for this state of affairs: (1) the presumption that REBs can handle all of these issues; or (2) failure to recognize of the broader ethical issues associated with research involving human subjects. Ultimately, we reject both of these explanations in favour of a third: key stakeholders lack the foresight and political will to make the necessary changes.

We begin by providing a moral geography of research involving human subjects. We describe the various stages involved in the planning and execution of research involving human subjects, and identify some of the ethical issues arising at each stage. We then explore the role of the REB within the process as a whole, contrasting the de jure role of the REB (as specified in the Tri-Council Policy Statement) with its de facto role in the current system. We conclude by calling for systematic efforts to redress the gaps in our current governance system.

We note, however, that though many of our points apply to all research involving human subjects, some are clearly more relevant to the conduct of health research in general and clinical research in particular. This is partly an artefact of the authors' expertise and partly because our Network is centred on health research. But the focus on health research also allows us to simplify our remarks since taking into account the sometimes significant variances within different types of research involving human subjects would be very difficult within the ambit of a single paper.

The problem: ethical tunnel vision

Ten years ago, one of us concluded a report on the governance of research involving human subjects for the Law Commission of Canada with the following observations (among others):

[W]ith respect to ethical conduct of HRIHS [health research involving human subjects] there [are] three central ethical objectives:

1. The promotion of socially beneficial research

2. The protection of research participants

3. As an overarching aim, the maintenance of trust between the research community and society as a whole.

   These objectives enjoy broad social endorsement and are represented in the numerous international, national and professional codes and [sic] aimed at the conduct of ethical research involving humans. But when we compare these three objectives with what actually takes place in the name of ethical research, we find a narrowing of concerns that could aptly be described as ethical tunnel vision, in which the three ethical objectives are given the most minimal instantiation. In effect, our current governance processes for HRIHS reduce research ethics to a dangerously simplistic concern for REB approval that is often functionally an approval of consent forms. The result is that the REB approval process and informed consent bear far more moral weight than they can possibly sustain. (2)

   Though there has been some progress toward improving this situation over the last decade, these remarks remain all too true today. For the most part, governments, sponsors, institutions, and researchers continue to view research ethics almost entirely through the lens of REB review.

   This situation is problematic for a number of reasons. First, even in ideal circumstances the REB system cannot meet the above stated ethical objectives on its own. Second, current circumstances are far from ideal. REB function is hampered by a chronic lack of commitment and resources at the macro (societal), meso (institutional), and micro (individual, i.e., project or researcher) levels.

   The narrow vision of research ethics as REB review, combined with the de facto limitations of the REB system, creates conditions ripe for institutional failure. First, tensions between REBs and researchers are exacerbated: REBs are overburdened by obligations (ethical and administrative) that should be borne by other stakeholders (e.g., governments, sponsors, institutions, and researchers); and the offloading of ethics onto the REB reinforces both the impression of REBs as the "ethics police" and the supposition that ethics is "somebody else's problem" (i.e., the REBs). Institutional failure, finally, threatens to undermine the trust between research community and society that is required for the on-going support and conduct of socially beneficial research. Recognizing the shortcomings in our ethics governance system is a first step toward managing the substantial risks involved.

   The Health Research Involving Human Subjects lifecycle

   In order to better understand the limitations of our current governance system, we need a clear picture of the process the system is supposed to govern; the process of, or various processes involved in, conducting health research involving human subjects (HRIHS). We call this the 'HRIHS lifecycle' (see Figure 1). The lifecycle includes twelve elements. Each element comprises an essential stage in the conduct of HRIHS, and each occurs at one or more levels (macro, meso, or micro). Furthermore, the ethical issues arising at each stage are themselves macro, meso, micro, or some combination thereof. It is important to distinguish the level of the element from the level at which the ethical issue(s) arise because, in some cases, the level of the issue does not correspond to the level of the element. For example, though data analysis typically occurs at the micro level because it is the responsibility of the researcher(s), it raises important ethical issues at the micro, meso, and macro levels. The elements comprising the HRIHS lifecycle are: (1) priority setting; (2) education - scientific and ethical; (3) protocol design; (4) funding review; (5) ethics review; (6) recruitment; (7) informed consent; (8) monitoring; (9) study termination; (10) data analysis; (11) knowledge transfer (KT); and, (12) quality assurance and quality improvement (QA/QI) -of all relevant processes.

   [FIGURE 1 OMITTED]

   For the most part, the elements themselves are self-explanatory. However, our decision to focus on these particular elements and to represent them in this manner is not. Clearly the HRIHS lifecycle could include more or less elements, with corresponding levels of detail. Furthermore, the process as a whole could be represented differently. For example, we could have represented the elements non-sequentially or linearly as opposed to cyclically. We hope that the HRIHS lifecycle - as represented here - strikes an effective balance between comprehensiveness and economy, but our primary aim is to bring out the procedural and ultimately ethical complexity of HRIHS. We have represented the whole process cyclically instead of linearly in order to emphasize the importance of continuous learning in the HRIHS enterprise; the results of today's research should inform the research of tomorrow. But we recognize that this picture is highly idealized. In reality, the elements are not always sequential (e.g., recruitment and consent can occur after ongoing review by a data and safety monitoring board (DSMB) has begun; and opportunities for KT arise at multiple points in the lifecycle). We also acknowledge that in some cases it may be misleading to represent some elements as discrete steps (e.g., REBs are responsible for reviewing protocol design, and the responsibilities of a DSMB include recommending study termination). Furthermore, it is far from clear that the real world of HRIHS is accurately represented as a closed feedback loop given the fragmentary character of the enterprise and the lack of a systematic outcomes measurement or QA. (3) Finally, as noted above, life cycles for different areas of health research (e.g., research in population health, health behaviour, and infectious disease) would require somewhat different diagrams even though they would share many common elements with those indicated in Figure 1. Hence, Figure 1 should be viewed as an idealized picture of the HRIHS lifecycle, not a literal description.

   The elements and the ethical issues

   The crucial goal, for present purposes, is the illumination of the ethical complexity and richness of HRIHS. One of the things that the HRIHS lifecycle makes clear is that ethics review comprises but one of the twelve elements involved in the conduct of HRIHS. The other eleven elements are both procedurally and ethically distinct, raising unique challenges of their own. Since we will discuss the role of the REB in detail in later sections, we focus here on the other eleven elements in the HRIHS lifecycle. Assuming the above stated ethical objectives of research governance (i.e., the promotion of socially beneficial research, the protection of research participants, and the maintenance of trust) we highlight key macro, meso, and micro ethical issues arising at each step. We also indicate the level at which each element occurs. The results of this analysis are summarized in Table 1.

   Priority setting

   Level: Priority setting generally occurs at the macro and meso levels.

   Ethical issues: Health research requires funding and, though the private sector funds an increasingly large share of health research, the public sector remains a significant source of financial support, particularly for "curiosity-driven" research. Funding priorities include...